Appropriate definition of childbirth using Japanese administrative database: A cross-sectional cohort validation study

Appropriate definition of childbirth using Japanese administrative database: A cross-sectional cohort validation study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Appropriate definition of childbirth using Japanese administrative database: A cross-sectional cohort validation study Miyuki Koizumi, Hiroki Nakajima, Yuichi Nishioka, Emiri Morita, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8284846/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Claims data analysis is useful in clinical research. However, no validation studies have been conducted using established algorithms to define childbirth among women. The aim of this study was to establish and validate algorithms to define childbirth from a claims database. Methods The DeSC database, including claims data for approximately 13 million people, as well as parent–child identifiers (IDs) as family information obtained from insurers, was used. Seven algorithms were designed using combinations of diagnosis-related codes with a suspected flag for childbirth (A), diagnosis-related codes without a suspected flag (B), and medical procedure codes (C). The combinations were A, B, C, A and/or C, and B and/or C. Parent–child IDs were used to determine the mother’s month and year of childbirth based on the child’s month and year of birth. The gold standard for the month and year of childbirth was defined as the child’s month and year of birth among women aged 15–49 years linked by parent–child IDs during the observation period. We calculated sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Kappa Index, and Youden Index for each algorithm. To validate algorithms for estimating second childbirth during the observation period, which would become useful and valuable in defining childbirth, identification of second childbirth was started 2–24 months after the first, given that the average age difference was two years. Results A total of 854,626 women were included in this study, of whom 37,934 were aged 15–49 years at the time of parent–child ID assignment and classified as experiencing childbirth during the observation period. The algorithm with the highest value was “A or C” (Kappa Index: 0.69, sensitivity: 65.8%, specificity: 99.0%, PPV: 74.4%, and NPV: 98.4%). For second childbirth, algorithm “A or C” showed that 11-month difference had the highest Youden Index at 0.551. Conclusion We developed algorithms based on claims data and established an optimal algorithm for estimating childbirth. This validated algorithm can be used for accurate estimation of childbirth to clarify pregnancy- and childbirth-related diseases in future claims database studies. Statistical Epidemiology Claims data Childbirth Validation Background Administrative claims databases provide longitudinal data, including medical inpatient claims, medical outpatient claims, medication use, and major procedures. In recent years, the number of epidemiological studies conducted using claims data has increased significantly because they are extremely useful and reflect real-world data with minimal bias. Claims data are not originally for study purposes but for obtaining medical reimbursement; however, there is insufficient clinical information regarding diagnoses, examination results, and disease severity [ 1 ]. Therefore, verifying the validity of the information recorded in claims data is essential [ 2 ]. Various diseases and conditions such as diabetes, cardiovascular diseases, and death have been defined using claims databases, and validation studies have been conducted [ 3 , 4 ]. However, validation studies regarding the definition of diseases and conditions, which are important for accuracy, remain limited. Pregnancy and the postpartum period are expected to be among the most important fields for studies conducted using real-world data effectively. Studies conducted using claims data related to pregnancy, childbirth, and the postpartum period have revealed various issues such as infant maltreatment, pertussis vaccination rates among parents with infants, prenatal care, postpartum depression, pregnancy- and lactation-associated osteoporosis, and use of antiepileptic drugs in pregnancy [ 5 – 10 ]. In these studies, date of childbirth was identified using claims data for medical procedure or diagnosis-related codes regarding childbirth, hospitalisations for childbirth, and birth certificate records [ 5 – 8 ]. Algorithms based on claims data regarding pregnancy period, date of childbirth, and pregnancy outcome have been developed in countries such as the United States, Germany, France, and Japan. The validity of these algorithms has been verified using randomly sampled electronic medical and birth certificate records [ 11 – 15 ]. However, it is necessary to validate them in a group that includes women without a history of childbirth, as the actual claims database includes many women without a history of childbirth during the observation period. Therefore, it is important to define childbirth using large-scale claims data among women in general. The aim of this study was to establish an algorithm for identifying the month and year of childbirth from a claims database, targeting women regardless of their childbirth history. We used a claims database that includes family information obtained from insurers and verified the validity of the algorithm in defining childbirth based on the claims data, using the childbirth month and year calculated from the family information as the gold standard. This algorithm is also expected to be applicable to other claims databases. Methods Study design A cross-sectional observational analysis was performed using the DeSC database (DeSC Healthcare, Tokyo, Japan). The database comprised data of women whose husbands were identified from family information among those whose data were registered between 2014 and 2023. This study was approved by the Ethics Committee of Nara Medical University (approval no. 1123) and conducted in accordance with the principles outlined in the Declaration of Helsinki. All patient data were anonymised prior to analysis. The requirement for informed consent was waived because of the retrospective nature of the study. Childbirth algorithm In previous studies conducted using claims data, Kasahara et al. and Wang et al. identified childbirth based on diagnosis-related or medical procedure codes [ 9 , 16 ], and Kim et al. and Park et al. identified childbirth based on medical procedure codes only[ 7 , 17 ]. These codes were extracted and modified by expert Obstetricians in Nara Medical University to define childbirth and combined to create algorithms. For disease names just before delivery or during delivery and childbirth, the codes that appeared during delivery and childbirth were selected. The specific diagnosis-related codes are listed in Table S3, and codes corresponding to O14–15, O32–34, O42, O45, O48, O60–69, O70–72, O74–75, O80–84, O86–87, O90, and O99 in the International Classification of Diseases, Tenth Revision (ICD-10) were selected. For medical procedure codes, those observed during childbirth, such as labour induction, vacuum extraction, forceps delivery, and caesarean section (Table S5) were selected. Seven algorithms for the claims-based definition of childbirth were designed. Algorithms 1 –7 were constructed using a combination of three elements: (A) whether they had diagnosis-related codes regarding childbirth with a suspected flag, (B) whether they had diagnosis-related codes regarding childbirth without a suspected flag, and (C) whether they had medical procedure codes regarding childbirth as A, B, C, A and/or C, B and/or C (Table 1 , S3, and S5). The rate of preterm births in Japan in 2022 was approximately 5.6% [ 18 ]. Additionally, previous reports have shown that the concordance rate between the algorithm and the gold standard is relatively low for preterm births [ 11 ]. Data of women with preterm births might influence analysis results; therefore, algorithms 8–13 for childbirths excluding potentially preterm births were set up. “Diagnosis-related codes regarding childbirths excluding potentially preterm births” such as premature rupture of membranes and obstructed labour due to malposition, and “diagnosis-related codes regarding childbirths as preterm births” such as placental abruption and medically indicated preterm birth, were determined. Algorithms 8–13 were also constructed by combining these three elements: A, B, C, A and/or C, and B and/or C (Table S2, S4, and S5). The latest month and year recorded in the claims data for diagnosis-related codes or medical procedure codes regarding childbirth were defined as the month and year of childbirth. In the algorithms used in this study, women with only one childbirth during the observation period were identified as the childbirth, and women with multiple childbirths during the observation period were identified as the last childbirth. The month of discharge may follow the month of childbirth, and the month of childbirth and that of diagnosis-related codes or medical procedure codes may differ by 1 month. Therefore, “the month and year of childbirth from algorithms or 1 month before the month and year of childbirth from algorithms” and “the month and year of childbirth from the parent–child identifier (ID)” were defined as the same. Data sources The DeSC database includes data on medical insurance for the elderly in the later stage of life, health insurance association set up by a company fulfilling certain conditions for private and government employees established with the approval of the Japanese government, and national health insurance for self-employed people and freelancers. This database was constructed using monthly claims from all medical institutions and pharmacies, specific health checkups, and registries in Japan submitted between April 2014 and January 2023, including approximately 13 million insured people (approximately 10.5% of the Japanese population). The DeSC database provides information on encrypted personal identifiers, month and year of birth, sex, ICD-10 codes, medicine codes (name, dose, and administration period), and medical procedure codes for insured people. Moreover, the DeSC database includes parent–child IDs as family information that indicate parent–child relationships, obtained from consenting insurers. A child who has enrolled in the same insurance system as the mother and/or father is linked to the mother and/or father when the parent has a parent–child ID. However, in the case of adopted children, parent–child IDs are not assigned. Linking the child's month and year of birth with the mother using parent–child IDs enabled the identification of the mother's month and year of childbirth. The gold standard for month and year of childbirth The gold standard for month and year of childbirth was defined as the child’s month and year of birth among women aged 15–49 years who were linked by parent–child IDs during the observation period. The insurers had records of insured people, which led to an accurate understanding of family relationships. The provision of this information regarding family relationships was strictly reviewed from the perspective of personal data protection. The provision of data from the DeSC database was permitted because the gold standard information was indispensable for the execution of this validation study. The rate of childbirth among women aged 15–49 in Japan in 2022 was 99.9% [ 18 ]. Therefore, age in this study was limited to 15–49 years to make the determined date of childbirth from the parent–child ID more accurate. Statistical analysis Identifying childbirth The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Kappa Index, and Youden Index for each algorithm with corresponding 95% confidence intervals were computed. Sensitivity and specificity were calculated by identifying whether the month and year of giving birth (-1 ~ 0 month) defined by the algorithms coincided with that of childbirth from the parent-–child ID. PPV was the proportion of women identified by an algorithm as experiencing childbirth who had truly given birth. NPV was the proportion of women identified by an algorithm as not experiencing childbirth, who had not truly given birth. The Kappa Index was calculated for the agreement between each algorithm and the gold standard to identify the algorithms that maximised Kappa [ 19 ]. The Youden Index was calculated as (sensitivity + specificity)-1, to weigh sensitivity and specificity equally. Identifying second childbirth during the observation period Diagnosis-related codes regarding childbirth may be used repeatedly for prepartum and postpartum visits. Diagnosis-related and medical procedure codes for the same childbirth may appear in different months. In identifying multiple childbirths in the database, there is a risk of misrepresenting single childbirths as multiple childbirths within a short period. The number of months that passed after the consecutive diagnosis-related codes or medical procedure codes may indicate another childbirth, rather than the same one. To confirm whether women with parent–child IDs had more than two childbirths using algorithms during the observation period, the validity of the difference in months based on each algorithm was verified: each month from 2 to 24 months after the first childbirth, given that the average age difference was reported to be two years in Japan in 2022 [ 18 ]. The Youden Index, sensitivity, specificity, PPV, NPV, and kappa index for each algorithm in each month for second childbirth were calculated. All analyses were performed using the Microsoft SQL Server 2022 Standard (Microsoft Corp., Redmond, WA, USA) and IBM SPSS for Windows (version 29.0; IBM Corp., Armonk, NY, USA). Results Participant characteristics In total, 854,626 women with husbands were identified in the DeSC database based on family information from the parent–child IDs. Table 2 presents the characteristics of the women included in this study. At the time of parent–child ID assignment, 37,934 women aged 15 to 49 (validation cohort birth rate, 4.4%) were classified as experiencing childbirth during the observation period (Table S1). Sixty-eight registered women with a parent-child ID were ineligible, as they did not fall within the target registrant age range of 15-49 years (0.2% of women with parent–child IDs). Validating algorithms [1 – 7] for childbirth Table 3 presents the accuracy of the administrative data algorithms in identifying women with childbirth experience and the month and year of their deliveries based on the DeSC database. The accuracy assessment of Algorithm 4, using diagnosis-related codes with a suspected flag or medical procedure codes regarding childbirth, showed the highest value, with a sensitivity of 65.8%, specificity of 99.0%, PPV of 74.4%, NPV of 98.4%, kappa index of 0.69, and Youden Index of 0.65, respectively. Algorithm 6, which was created using diagnosis-related codes or medical procedure codes, resulted in a sensitivity of 63.9%, specificity of 99.0%, PPV of 74.7%, and Kappa Index of 0.68. Validating algorithms [8 – 13] for identifying childbirth, excluding potentially preterm birth Table S6 shows the accuracy of the administrative data algorithms in identifying women with childbirth experience and the month and year of their deliveries excluding those with a high possibility of preterm birth. Algorithm 11, which was created using diagnosis-related codes with a suspected flag or medical procedure code, showed the highest value, with a sensitivity of 65.3%, specificity of 99.0%, PPV of 74.8%, NPV of 98.4%, Kappa Index of 0.68, and Youden Index of 0.64, respectively. Algorithm 13, which was created using diagnosis-related codes or medical procedure codes, resulted in a sensitivity of 63.6%, specificity of 99.0%, PPVs of 75.0%, and Kappa Index of 0.68, respectively. Validating algorithms for identifying second childbirth Table S7 shows the accuracy of the administrative data algorithms in identifying second childbirth when the evaluation of the algorithms began 2–24 months after the first childbirth during the observation period. When the observation of diagnosis-related codes began 11 months after the first childbirth in Algorithm 4, the Youden Index was 0.5519, the highest value, and the sensitivity, specificity, PPV, NPV, and Kappa Index were 57.7%, 97.5%, 80.8%, 92.6%, and 0.62, respectively. Discussion In this study, we established and validated algorithms for defining the month and year of childbirth based on a claims database among women, regardless of their childbirth history. Algorithm 4 involving diagnosis-related codes with a suspected flag or medical procedure codes regarding childbirth yielded the highest value, with a Kappa Index of 0.69 and a sensitivity of 65.8% for all childbirths. The PPV was 74.4% in all algorithms, with the highest value being 77.0%, showing minimal difference. Algorithm 11 excluded births with a high possibility of being preterm from Algorithm 4. Algorithms 4 and 11 showed a nearly similar Youden Index at 0.648 and 0.643, respectively. Considering second childbirth using the claims database, at the start of observation, 11 months after the first childbirth, algorithm 4 yielded the highest Youden Index at 0.551. Although sensitivity was not high at 57.7%, specificity, PPV, and NPV were high at 97.5%, 80.8%, and 92.6%, respectively. Algorithms 1–7 yielded PPV and NPV of approximately 75% and 95%, respectively, for all childbirths. Algorithm 4, with the highest Kappa Index at 0.69, was the most accurate for identifying childbirth based on claims data. Comparing algorithms 4 and 6, using diagnosis-related codes with a suspected flag yielded a higher sensitivity than without a suspected flag, and both categories had an almost similar specificity. The PPVs of algorithms 4 and 6 were almost the same because there were very few cases of diagnosis-related codes with a suspected flag regarding childbirth among non-pregnant women. Moreover, we established and validated an algorithm for estimating the date of second childbirth. In the analysis, 11 months after the first childbirth, algorithm 4 yielded the highest Youden Index at 0.5519. Women who do not breastfeed after childbirth can ovulate and become pregnant 6 weeks after childbirth [10]. Considering that a baby is born at full term in 37–41 weeks, it takes approximately 11 months to give birth again and at least 6 weeks to become pregnant again, which may explain why the 11-month Youden Index was the highest. In previous studies, the identification of date of childbirth varied based on factors such as childbirth-related claims data, hospitalisations for childbirth, and birth certificate records, and the validity of algorithm for defining childbirth were based on electronic medical records and birth certificate records [5-8, 11-14]. All these studies were conducted only among women who had given birth. In a previous study conducted in Japan, the algorithm was validated using medical claims data of obstetric patients at a single institute. This previous validation study included only women who had given birth and was conducted at a high-level medical institution [15]. Under this condition, it is possible that there was a bias, such as with high-risk pregnancies. All these previous studies were conducted among women who had given birth, and there have been no validation studies on childbirth with established algorithms for defining date of childbirth, regardless of whether they had given birth or not. Therefore, our study targeting women in the general population, with a less-biased claims database, is significant. In this study, the proportions of mothers aged 15–19, 20–24, 25–29, 30–39, 40–44, and 45–49 years at the time of childbirth were 1.0%, 8.6%, 24.1%, 34.7%, 24.7%, 6.6%, and 0.3%, respectively. The proportions of mothers aged 15–19, 20–24, 25–29, 30–39, 40–44, and 45–49 years at the time of childbirth in Japan in 2022 were 0.6%, 6.9%, 26.3%, 36.3%, 23.8%, 6.0%, and 0.2%, respectively [18]. The proportions reported in our study and Japan were nearly similar; therefore, the inclusion of mothers in this study was thought to be less biased. The main strength of this study is that it is the first validation study of date of childbirth in a large population of over 850,000 women regardless of birth status. The algorithm based on the claims data in this study would be widely useable and applicable to other claims databases that do not contain the parent–child information that identified childbirth. ICD-10 codes were used in this study; therefore, this algorithm is applicable globally. This study has several limitations. First, the validation analysis only included women with the parent–child IDs as family information. However, the proportion of women who gave birth was almost the same as the national average; therefore, bias was considered small. Second, women with normal deliveries could not be evaluated by the algorithms in this study because medical insurance is generally not used for normal deliveries. The algorithms used in this study cannot identify childbirths without medical procedures. Therefore, when targeting normal deliveries that do not involve medical procedure codes or diagnosis-related codes regarding childbirth, it is necessary to use administrative databases such as parent–child IDs. Third, there may have been a misallocation of parent–child IDs when linking parent-child relationships in the administrative database. However, the possibility of a link error was considered extremely low because four women aged under 15 (0.02%) and 64 women aged over 50 (0.01%) years were identified as mothers by parent–child IDs. Conclusions In this study, we established an optimal algorithm for estimating childbirth based on claims data. This validated algorithm will be useful for accurate estimation of childbirth in future claims database studies. Abbreviations ICD-10, International Classification of Diseases, Tenth Revision; ID, identifier; TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; PPV, positive predictive value; NPV, negative predictive value; 95% CI, 95 percent confidence interval Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of Nara Medical University (approval no. 1123), and conducted in accordance with the principles outlined in the Declaration of Helsinki. All patient data were anonymised prior to analysis. The requirement for informed consent was waived because of the retrospective nature of the study. Consent for publication Not applicable. Availability of data and materials The data that support the findings of this study are available from DeSC Healthcare, Inc., but restrictions apply to the availability of these data, which were used under the licence for the current study and are not publicly available. However, the data are available from the authors upon reasonable request and with permission from DeSC Healthcare, Inc. Competing interests YN received speaker fees from Novo Nordisk, Sanofi, Daiichi Sankyo, and DeSC Healthcare. YT received consultant fees from Novo Nordisk, Otsuka, and Recordati, and speaker fees from Novo Nordisk, Sumitomo Dainippon, Eli Lilly, Ono, Novartis, Nippon Boehringer Ingelheim, AstraZeneca, and Kyowa Kirin. The other authors declare no competing interests. Funding This study was supported by a grant from the Japan Ministry of Education, Culture, Sports, Science and Technology. (grant numbers: 22H03355, 23K24613, 24K22341, and 23K16362). Authors’ contributions M.K. designed the study, analyzed the data, and wrote the manuscript. H.N. provided advice on this paper and discussed the validity of statistical methods and specific methodologies. Y.N. provided advice on this paper and technical support in statistical methods and extraction of data. Y.T. supervised the study and reviewed the manuscript. E.M., T.N., T.M., and T.I. contributed to writing the manuscript. H.N. and Y.N. contributed equally to this study. Acknowledgements The DeSC database was provided by DeSC Healthcare, Inc. Under their academic research support program. References Hirose N, Ishimaru M, Morita K, Yasunaga H (2020) A review of studies using the Japanese National Database of Health Insurance Claims and Specific Health Checkups. Annals Clin Epidemiol 2:13–26 Hall GC, Sauer B, Bourke A, Brown JS, Reynolds MW, LoCasale R (2012) Guidelines for good database selection and use in pharmacoepidemiology research. Pharmacoepidemiol Drug Saf 21:1–10 Ito F, Togashi S, Sato Y, Masukawa K, Sato K, Nakayama M et al (2023) Validation study on definition of cause of death in Japanese claims data. PLoS ONE 18:e0283209 Fujihara K, Yamada-Harada M, Matsubayashi Y, Kitazawa M, Yamamoto M, Yaguchi Y et al (2021) Accuracy of Japanese claims data in identifying diabetes-related complications. Pharmacoepidemiol Drug Saf 30:594–601 Roberts SCM, Schulte A, Zaugg C, Leslie DL, Corr TE, Liu G (2023) Association of pregnancy-specific alcohol policies with infant morbidities and maltreatment. JAMA Netw Open 6:e2327138 Marchal C, Belhassen M, Guiso N, Jacoud F, Cohen R, Le Pannerer M et al (2022) Cocooning strategy: pertussis vaccination coverage rate of parents with a new-born in 2016 and 2017 in France. Front Pediatr 10:988674 Kim S, Kim C, Kim JH (2024) Antenatal care inequalities in South Korea: an analysis of health insurance claims data (2013–2022) in a high-resource, high-use country. Int J Gynaecol Obstet 166:718–726 Steenland MW, Trivedi AN (2023) Association of Medicaid expansion with postpartum depression treatment in Arkansas. JAMA Health Forum 4:e225603 Kasahara K, Tanaka-Mizuno S, Tsuji S, Ohashi M, Kasahara M, Kawasaki T et al (2024) Pregnancy and lactation-associated osteoporosis as a major type of premenopausal osteoporosis: a retrospective cohort study based on real-world data. BMC Pregnancy Childbirth 24:301 Ishikawa T, Obara T, Jin K, Nishigori H, Miyakoda K, Suzuka M et al (2019) Examination of the prescription of antiepileptic drugs to prenatal and postpartum women in Japan from a health administrative database. Pharmacoepidemiol Drug Saf 28:804–811 Zhu Y, Hampp C, Wang X, Albogami Y, Wei YJJ, Brumback BA et al (2020) Validation of algorithms to estimate gestational age at birth in the Medicaid Analytic eXtract-Quantifying the misclassification of maternal drug exposure during pregnancy. Pharmacoepidemiol Drug Saf 29:1414–1422 Moll K, Wong HL, Fingar K, Hobbi S, Sheng M, Burrell TA et al (2021) Validating claims-based algorithms determining pregnancy outcomes and gestational age using a linked claims-electronic medical record database. Drug Saf 44:1151–1164 Wentzell N, Schink T, Haug U, Ulrich S, Niemeyer M, Mikolajczyk R (2018) Optimizing an algorithm for the identification and classification of pregnancy outcomes in German claims data. Pharmacoepidemiol Drug Saf 27:1005–1010 Blotière PO, Weill A, Dalichampt M, Billionnet C, Mezzarobba M, Raguideau F et al (2018) Development of an algorithm to identify pregnancy episodes and related outcomes in health care claims databases: an application to antiepileptic drug use in 4.9 million pregnant women in France. Pharmacoepidemiol Drug Saf 27:763–770 Tajima K, Ishikawa T, Noda A, Matsuzaki F, Morishita K, Inoue R et al (2022) Development and validation of claims-based algorithms to identify pregnancy based on data from a university hospital in Japan. Curr Med Res Opin 38:1651–1654 Wang X, Meisel Z, Kellom K, Whitaker J, Strane D, Chatterjee A et al (2023) Receipt and duration of buprenorphine treatment during pregnancy and postpartum periods in a national privately-insured cohort. Drug Alcohol Depend Rep 9:100206 Park SJ, Choi NK, Seo KH, Park KH, Woo SJ (2015) Retinal vein occlusion and pregnancy, pre-eclampsia, and eclampsia: the results from a nationwide, population-based study using the national claim database. PLoS ONE 10:e0120067 Vital statistics. Tokyo: Ministry of Health, Labor and welfare (2022) https://www.mhlw.go.jp/toukei/saikin/hw/jinkou/kakutei22/index.html . Accessed 19 Apr 2025 Kundel HL, Polansky M (2003) Measurement of observer agreement. Radiology 228:303–308 Tables Table 1. Algorithms for identifying childbirth Cndition A Diagnosis-related codes regarding childbirth with a suspected flag Cndition B Diagnosis-related codes regarding childbirth without a suspected flag Condition C Medical procedure codes regarding childbirth Algorithm Condition Algorithm1 A Algorithm2 B Algorithm3 A and C Algorithm4 A or C Algorithm5 B and C Algorithm6 B or C Algorithm7 C Table 2. Characteristics of women in the validation study Age Number Percentage 0～4 7,162 0.8% 5～9 7,201 0.8% 10～14 7,505 0.9% 15～19 10,278 1.2% 20～24 20,312 2.4% 25～29 30,870 3.6% 30～34 44,518 5.2% 35～39 53,633 6.3% 40～44 62,345 7.3% 45～49 68,259 8.0% 50～54 65,660 7.7% 55～59 80,542 9.4% 60～64 150,979 17.7% 65～69 193,158 22.6% 70～74 52,204 6.1% Total 854,626 100% Table 3. Accuracy of administrative data algorithms for identifying women with childbirth experience and the month and year of their deliveries Algorithm Algorithm1 Algorithm2 Algorithm3 Algorithm4 Algorithm5 Algorithm6 Algorithm7 TP 21,854 20,947 12,279 24,957 12,083 24,223 15,123 TN 809,177 809,613 813,011 808,126 813,073 808,500 811,960 FN 16,080 16,987 25,655 12,977 25,851 13,711 22,811 FP 7,515 7,079 3,681 8,566 3,619 8,192 4,732 Sensitivity(%) 57.6% 55.2% 32.4% 65.8% 31.9% 63.9% 39.9% 95% CI 57.1% 54.7% 31.9% 65.3% 31.4% 63.4% 39.4% 58.1% 55.7% 32.8% 66.3% 32.3% 64.3% 40.4% Specificity (%) 99.1% 99.1% 99.5% 99.0% 99.6% 99.0% 99.4% 95% CI 99.1% 99.1% 99.5% 98.9% 99.5% 99.0% 99.4% 99.1% 99.2% 99.6% 99.0% 99.6% 99.0% 99.4% PPV (%) 74.4% 74.7% 76.9% 74.4% 77.0% 74.7% 76.2% 95% CI 73.9% 74.2% 76.3% 74.0% 76.3% 74.3% 75.6% 74.9% 75.3% 77.6% 74.9% 77.6% 75.2% 76.8% NPV (%) 98.1% 97.9% 96.9% 98.4% 96.9% 98.3% 97.3% 95% CI 98.0% 97.9% 96.9% 98.4% 96.9% 98.3% 97.2% 98.1% 98.0% 97.0% 98.4% 97.0% 98.4% 97.3% Kappa 0.64 0.62 0.44 0.69 0.44 0.68 0.51 Youden 0.57 0.54 0.32 0.65 0.31 0.63 0.39 Abbreviations: TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; 95% CI, 95 percent confidence interval; PPV, positive predictive value; NPV, negative predictive value. Additional Declarations The authors declare no competing interests. Supplementary Files SupplementaryTable1.xlsx Additional file (.xlsx) Supplementary Table 1. Characteristics of women who gave birth during the observation period in the validation study SupplementaryTable2.xlsx Supplementary Table 2. Algorithms for identifying childbirths excluding potentially preterm births SupplementaryTable3.xlsx Supplementary Table 3. Diagnosis-related codes regarding childbirths SupplementaryTable4.xlsx Supplementary Table 4. Diagnosis-related codes regarding childbirths excluding potentially preterm births SupplementaryTable5.xlsx Supplementary Table 5. Medical procedure codes regarding childbirth SupplementaryTable6.xlsx Supplementary Table 6. Accuracy of algorithms identifying women with childbirth experience and the month and year of their deliveries, excluding those likely to have preterm births Abbreviations: TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; 95% CI, 95 percent confidence interval; PPV, positive predictive value; NPV, negative predictive value. SupplementaryTable7.xlsx Supplementary Table 7. Accuracy of algorithms identifying subsequent childbirths based on month difference observed as childbirth-related codes Abbreviations: TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; 95% CI, 95 percent confidence interval; PPV, positive predictive value; NPV, negative predictive value. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8284846","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":555646313,"identity":"417219c6-38e0-4ea1-93cb-2a6474a33622","order_by":0,"name":"Miyuki Koizumi","email":"","orcid":"https://orcid.org/0009-0004-0837-5365","institution":"Nara Medical University","correspondingAuthor":false,"prefix":"","firstName":"Miyuki","middleName":"","lastName":"Koizumi","suffix":""},{"id":555815100,"identity":"a303c621-a20e-4c91-9b4c-c64dcf9696c5","order_by":1,"name":"Hiroki Nakajima","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9klEQVRIiWNgGAWjYFACxgaGBAYGHhDzAIOBjRyY8YAoLWzMQC0FacZgLQlE2QbUwsDw4XBiA4iDT4u59OHWDQ8Y7sjwz+8/CHQYc/r8sMMPgbbYyek2YNdi2ZfYdiOB4RmPxDGQwwzYcjfeTjMAakk2NjuAXYvBGUaQlsM8DBAtPLkbZyeAtBxI3EZIizxEi0S64ez0D8RpMYBoMUiQl87Bb4tlD0iLwTMew2PJQJUGCYYbpHMKgAzcfjHnYX9280fFHXu5wwcff/jw57+8/Oz0zR8+VNjJ4fQ+hITKJsDZBtiVI0khGSjfgFv1KBgFo2AUjEwAAH67Y8hPaGk/AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-5899-0903","institution":"Nara Medical University","correspondingAuthor":true,"prefix":"","firstName":"Hiroki","middleName":"","lastName":"Nakajima","suffix":""},{"id":555815101,"identity":"886bb32f-8a8c-4c26-a07c-8ac146ca615f","order_by":2,"name":"Yuichi Nishioka","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYHACxgdQhgGUPkBQCzNMKfFa2CTQtBAABgd4j1UX1NyT121v3sDMw2Anz8B4Fr81Bgf40m7POFZsuO3MsQKglmTDBoZzCQS08Jjd5mFLYNx2I8f8Nw8DM1D5GfwuBGkp5vmXYL/t/hsDoC31xGlh5m1LSNx2gwek5TBhLZIHeIylefsSkredSStgnGNw3LCNkF/4DvAYfub5lmC77fjhDQxvKqrl+SUIhJjC/Qco7gRF0xm8OhjkGzCE+HvwaxkFo2AUjIIRBwAC60DqE5H1OAAAAABJRU5ErkJggg==","orcid":"","institution":"Nara Medical University","correspondingAuthor":true,"prefix":"","firstName":"Yuichi","middleName":"","lastName":"Nishioka","suffix":""},{"id":555815102,"identity":"20cae777-8201-4f3e-b34e-0eb026d85d20","order_by":3,"name":"Emiri Morita","email":"","orcid":"","institution":"Nara Medical University","correspondingAuthor":false,"prefix":"","firstName":"Emiri","middleName":"","lastName":"Morita","suffix":""},{"id":555815103,"identity":"36069574-8c2e-415c-9e2b-353f97bb4d1d","order_by":4,"name":"Tomoya Myojin","email":"","orcid":"","institution":"Hamamatsu University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tomoya","middleName":"","lastName":"Myojin","suffix":""},{"id":555815104,"identity":"43fbdb03-7507-4bec-9148-f0fd647412f8","order_by":5,"name":"Tatsuya Noda","email":"","orcid":"","institution":"Kansai Medical University","correspondingAuthor":false,"prefix":"","firstName":"Tatsuya","middleName":"","lastName":"Noda","suffix":""},{"id":555815105,"identity":"0f3608c2-bb8b-4286-a94b-b4724a9ff1cb","order_by":6,"name":"Tomoaki Imamura","email":"","orcid":"","institution":"Nara Medical University","correspondingAuthor":false,"prefix":"","firstName":"Tomoaki","middleName":"","lastName":"Imamura","suffix":""},{"id":555815106,"identity":"d0c1b81f-d77d-44e0-879f-220a8ea0db4e","order_by":7,"name":"Yutaka Takahashi","email":"","orcid":"","institution":"Nara Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yutaka","middleName":"","lastName":"Takahashi","suffix":""}],"badges":[],"createdAt":"2025-12-05 06:37:51","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":true,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-8284846/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8284846/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97982054,"identity":"66458d54-5833-496d-af6b-997db9970632","added_by":"auto","created_at":"2025-12-11 12:59:44","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":90243,"visible":true,"origin":"","legend":"","description":"","filename":"BMCPublicHealthValidationstudyTable20250422.docx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/feeaeacda716664d0d1954b2.docx"},{"id":98424570,"identity":"f9663c9f-1656-4c36-91ec-1af755a4b8b1","added_by":"auto","created_at":"2025-12-17 16:33:29","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":342,"visible":true,"origin":"","legend":"","description":"","filename":"rs8284846.json","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/4d9053d4a09aec6fcbd7d663.json"},{"id":98424285,"identity":"244c7764-78f6-488e-9ee6-d1592446623b","added_by":"auto","created_at":"2025-12-17 16:33:08","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":90848,"visible":true,"origin":"","legend":"","description":"","filename":"rs82848460enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/31bc2c2c51bcd9c4dac34a17.xml"},{"id":98423617,"identity":"d04788f6-0d3e-48ea-b0be-caaa9e2e46dc","added_by":"auto","created_at":"2025-12-17 16:32:26","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":88103,"visible":true,"origin":"","legend":"","description":"","filename":"rs82848460structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/40ae6c43a6acfe821eb3fc9e.xml"},{"id":97982059,"identity":"9b30d381-11a7-4d71-875f-2d7ae424b33f","added_by":"auto","created_at":"2025-12-11 12:59:45","extension":"html","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":97860,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/dfcb8cf6f16e938a22f3bed8.html"},{"id":98622139,"identity":"58c9a92d-8884-4197-9374-eff9b2fc46d4","added_by":"auto","created_at":"2025-12-19 16:46:20","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1242165,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/28f3c791-d477-4e83-a782-14a314becc45.pdf"},{"id":97982051,"identity":"7e87a1bd-9ef9-4004-82bd-35b7526b870f","added_by":"auto","created_at":"2025-12-11 12:59:44","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":10481,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional file (.xlsx)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Table 1. Characteristics of women who gave birth during the observation period in the validation study\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"SupplementaryTable1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/7dda5a13b41ad709996f7d60.xlsx"},{"id":97982052,"identity":"090e1bdd-7f8d-4dbc-9df2-40e0ad82f5a4","added_by":"auto","created_at":"2025-12-11 12:59:44","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":10566,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table 2. Algorithms for identifying childbirths excluding potentially preterm births\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"SupplementaryTable2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/c677dac754a743190949d7df.xlsx"},{"id":98422744,"identity":"4662e732-8b84-4044-9e12-b52c7e41ee54","added_by":"auto","created_at":"2025-12-17 16:31:25","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":15910,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table 3. Diagnosis-related codes regarding childbirths\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"SupplementaryTable3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/8992891f3614b4a1cebcf404.xlsx"},{"id":98423254,"identity":"d336ed2f-7195-4555-8965-5df10e1f7c46","added_by":"auto","created_at":"2025-12-17 16:32:01","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":15252,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table 4. Diagnosis-related codes regarding childbirths excluding potentially preterm births\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"SupplementaryTable4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/f4ffd73f907fc3902caae644.xlsx"},{"id":98423846,"identity":"2fdb9315-2496-4266-a0bc-faa5aaaa71b9","added_by":"auto","created_at":"2025-12-17 16:32:41","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":10139,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table 5. Medical procedure codes regarding childbirth\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"SupplementaryTable5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/8f3fc4e3c6bfe1fe12bbfc3c.xlsx"},{"id":97982061,"identity":"457d18c1-6842-4290-92a3-3e137e551fe0","added_by":"auto","created_at":"2025-12-11 12:59:45","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":13373,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table 6. Accuracy of algorithms identifying women with childbirth experience and the month and year of their deliveries, excluding those likely to have preterm births\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; 95% CI, 95 percent confidence interval; PPV, positive predictive value; NPV, negative predictive value.\u003c/p\u003e","description":"","filename":"SupplementaryTable6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/e655d7c7b206ec01fb25ad7b.xlsx"},{"id":98424036,"identity":"6ee91735-63d7-4daf-94e1-65f0e91199c2","added_by":"auto","created_at":"2025-12-17 16:32:52","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":81334,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table 7. Accuracy of algorithms identifying subsequent childbirths based on month difference observed as childbirth-related codes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAbbreviations: TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; 95% CI, 95 percent confidence interval; PPV, positive predictive value; NPV, negative predictive value.\u003c/p\u003e","description":"","filename":"SupplementaryTable7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8284846/v1/3516fae30d629ffb8a0edda1.xlsx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eAppropriate definition of childbirth using Japanese administrative database: A cross-sectional cohort validation study \u003c/p\u003e","fulltext":[{"header":"Background","content":"\u003cp\u003eAdministrative claims databases provide longitudinal data, including medical inpatient claims, medical outpatient claims, medication use, and major procedures. In recent years, the number of epidemiological studies conducted using claims data has increased significantly because they are extremely useful and reflect real-world data with minimal bias. Claims data are not originally for study purposes but for obtaining medical reimbursement; however, there is insufficient clinical information regarding diagnoses, examination results, and disease severity [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Therefore, verifying the validity of the information recorded in claims data is essential [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Various diseases and conditions such as diabetes, cardiovascular diseases, and death have been defined using claims databases, and validation studies have been conducted [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. However, validation studies regarding the definition of diseases and conditions, which are important for accuracy, remain limited.\u003c/p\u003e\u003cp\u003ePregnancy and the postpartum period are expected to be among the most important fields for studies conducted using real-world data effectively. Studies conducted using claims data related to pregnancy, childbirth, and the postpartum period have revealed various issues such as infant maltreatment, pertussis vaccination rates among parents with infants, prenatal care, postpartum depression, pregnancy- and lactation-associated osteoporosis, and use of antiepileptic drugs in pregnancy [\u003cspan additionalcitationids=\"CR6 CR7 CR8 CR9\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. In these studies, date of childbirth was identified using claims data for medical procedure or diagnosis-related codes regarding childbirth, hospitalisations for childbirth, and birth certificate records [\u003cspan additionalcitationids=\"CR6 CR7\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Algorithms based on claims data regarding pregnancy period, date of childbirth, and pregnancy outcome have been developed in countries such as the United States, Germany, France, and Japan. The validity of these algorithms has been verified using randomly sampled electronic medical and birth certificate records [\u003cspan additionalcitationids=\"CR12 CR13 CR14\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. However, it is necessary to validate them in a group that includes women without a history of childbirth, as the actual claims database includes many women without a history of childbirth during the observation period. Therefore, it is important to define childbirth using large-scale claims data among women in general.\u003c/p\u003e\u003cp\u003eThe aim of this study was to establish an algorithm for identifying the month and year of childbirth from a claims database, targeting women regardless of their childbirth history. We used a claims database that includes family information obtained from insurers and verified the validity of the algorithm in defining childbirth based on the claims data, using the childbirth month and year calculated from the family information as the gold standard. This algorithm is also expected to be applicable to other claims databases.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eStudy design\u003c/h2\u003e\n \u003cp\u003eA cross-sectional observational analysis was performed using the DeSC database (DeSC Healthcare, Tokyo, Japan). The database comprised data of women whose husbands were identified from family information among those whose data were registered between 2014 and 2023.\u003c/p\u003e\n \u003cp\u003eThis study was approved by the Ethics Committee of Nara Medical University (approval no. 1123) and conducted in accordance with the principles outlined in the Declaration of Helsinki. All patient data were anonymised prior to analysis. The requirement for informed consent was waived because of the retrospective nature of the study.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eChildbirth algorithm\u003c/h3\u003e\n\u003cp\u003eIn previous studies conducted using claims data, Kasahara et al. and Wang et al. identified childbirth based on diagnosis-related or medical procedure codes [\u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e], and Kim et al. and Park et al. identified childbirth based on medical procedure codes only[\u003cspan class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e]. These codes were extracted and modified by expert Obstetricians in Nara Medical University to define childbirth and combined to create algorithms. For disease names just before delivery or during delivery and childbirth, the codes that appeared during delivery and childbirth were selected. The specific diagnosis-related codes are listed in Table S3, and codes corresponding to O14\u0026ndash;15, O32\u0026ndash;34, O42, O45, O48, O60\u0026ndash;69, O70\u0026ndash;72, O74\u0026ndash;75, O80\u0026ndash;84, O86\u0026ndash;87, O90, and O99 in the International Classification of Diseases, Tenth Revision (ICD-10) were selected. For medical procedure codes, those observed during childbirth, such as labour induction, vacuum extraction, forceps delivery, and caesarean section (Table S5) were selected.\u003c/p\u003e\n\u003cp\u003eSeven algorithms for the claims-based definition of childbirth were designed. Algorithms \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026ndash;7 were constructed using a combination of three elements: (A) whether they had diagnosis-related codes regarding childbirth with a suspected flag, (B) whether they had diagnosis-related codes regarding childbirth without a suspected flag, and (C) whether they had medical procedure codes regarding childbirth as A, B, C, A and/or C, B and/or C (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, S3, and S5). The rate of preterm births in Japan in 2022 was approximately 5.6% [\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e]. Additionally, previous reports have shown that the concordance rate between the algorithm and the gold standard is relatively low for preterm births [\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e]. Data of women with preterm births might influence analysis results; therefore, algorithms 8\u0026ndash;13 for childbirths excluding potentially preterm births were set up. \u0026ldquo;Diagnosis-related codes regarding childbirths excluding potentially preterm births\u0026rdquo; such as premature rupture of membranes and obstructed labour due to malposition, and \u0026ldquo;diagnosis-related codes regarding childbirths as preterm births\u0026rdquo; such as placental abruption and medically indicated preterm birth, were determined. Algorithms 8\u0026ndash;13 were also constructed by combining these three elements: A, B, C, A and/or C, and B and/or C (Table S2, S4, and S5).\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eThe latest month and year recorded in the claims data for diagnosis-related codes or medical procedure codes regarding childbirth were defined as the month and year of childbirth. In the algorithms used in this study, women with only one childbirth during the observation period were identified as the childbirth, and women with multiple childbirths during the observation period were identified as the last childbirth. The month of discharge may follow the month of childbirth, and the month of childbirth and that of diagnosis-related codes or medical procedure codes may differ by 1 month. Therefore, \u0026ldquo;the month and year of childbirth from algorithms or 1 month before the month and year of childbirth from algorithms\u0026rdquo; and \u0026ldquo;the month and year of childbirth from the parent\u0026ndash;child identifier (ID)\u0026rdquo; were defined as the same.\u003c/p\u003e\n\u003ch3\u003eData sources\u003c/h3\u003e\n\u003cp\u003eThe DeSC database includes data on medical insurance for the elderly in the later stage of life, health insurance association set up by a company fulfilling certain conditions for private and government employees established with the approval of the Japanese government, and national health insurance for self-employed people and freelancers. This database was constructed using monthly claims from all medical institutions and pharmacies, specific health checkups, and registries in Japan submitted between April 2014 and January 2023, including approximately 13\u0026nbsp;million insured people (approximately 10.5% of the Japanese population). The DeSC database provides information on encrypted personal identifiers, month and year of birth, sex, ICD-10 codes, medicine codes (name, dose, and administration period), and medical procedure codes for insured people. Moreover, the DeSC database includes parent\u0026ndash;child IDs as family information that indicate parent\u0026ndash;child relationships, obtained from consenting insurers. A child who has enrolled in the same insurance system as the mother and/or father is linked to the mother and/or father when the parent has a parent\u0026ndash;child ID. However, in the case of adopted children, parent\u0026ndash;child IDs are not assigned. Linking the child\u0026apos;s month and year of birth with the mother using parent\u0026ndash;child IDs enabled the identification of the mother\u0026apos;s month and year of childbirth.\u003c/p\u003e\n\u003ch3\u003eThe gold standard for month and year of childbirth\u003c/h3\u003e\n\u003cp\u003eThe gold standard for month and year of childbirth was defined as the child\u0026rsquo;s month and year of birth among women aged 15\u0026ndash;49 years who were linked by parent\u0026ndash;child IDs during the observation period. The insurers had records of insured people, which led to an accurate understanding of family relationships. The provision of this information regarding family relationships was strictly reviewed from the perspective of personal data protection. The provision of data from the DeSC database was permitted because the gold standard information was indispensable for the execution of this validation study. The rate of childbirth among women aged 15\u0026ndash;49 in Japan in 2022 was 99.9% [\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e]. Therefore, age in this study was limited to 15\u0026ndash;49 years to make the determined date of childbirth from the parent\u0026ndash;child ID more accurate.\u003c/p\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003eStatistical analysis\u003c/h2\u003e\n \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\n \u003ch2\u003eIdentifying childbirth\u003c/h2\u003e\n \u003cp\u003eThe sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Kappa Index, and Youden Index for each algorithm with corresponding 95% confidence intervals were computed. Sensitivity and specificity were calculated by identifying whether the month and year of giving birth (-1\u0026thinsp;~\u0026thinsp;0 month) defined by the algorithms coincided with that of childbirth from the parent-\u0026ndash;child ID. PPV was the proportion of women identified by an algorithm as experiencing childbirth who had truly given birth. NPV was the proportion of women identified by an algorithm as not experiencing childbirth, who had not truly given birth. The Kappa Index was calculated for the agreement between each algorithm and the gold standard to identify the algorithms that maximised Kappa [\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e]. The Youden Index was calculated as (sensitivity\u0026thinsp;+\u0026thinsp;specificity)-1, to weigh sensitivity and specificity equally.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003ch3\u003eIdentifying second childbirth during the observation period\u003c/h3\u003e\n\u003cp\u003eDiagnosis-related codes regarding childbirth may be used repeatedly for prepartum and postpartum visits. Diagnosis-related and medical procedure codes for the same childbirth may appear in different months. In identifying multiple childbirths in the database, there is a risk of misrepresenting single childbirths as multiple childbirths within a short period. The number of months that passed after the consecutive diagnosis-related codes or medical procedure codes may indicate another childbirth, rather than the same one. To confirm whether women with parent\u0026ndash;child IDs had more than two childbirths using algorithms during the observation period, the validity of the difference in months based on each algorithm was verified: each month from 2 to 24 months after the first childbirth, given that the average age difference was reported to be two years in Japan in 2022 [\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e]. The Youden Index, sensitivity, specificity, PPV, NPV, and kappa index for each algorithm in each month for second childbirth were calculated. All analyses were performed using the Microsoft SQL Server 2022 Standard (Microsoft Corp., Redmond, WA, USA) and IBM SPSS for Windows (version 29.0; IBM Corp., Armonk, NY, USA).\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eParticipant characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn total, 854,626 women with husbands were identified in the DeSC database based on family information from the parent–child IDs. Table 2 presents the characteristics of the women included in this study. At the time of parent–child ID assignment, 37,934 women aged 15 to 49 (validation cohort birth rate, 4.4%) were classified as experiencing childbirth during the observation period (Table S1). Sixty-eight registered women with a parent-child ID were ineligible, as they did not fall within the target registrant age range of 15-49 years (0.2% of women with parent–child IDs).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidating algorithms [1\u003c/strong\u003e–\u003cstrong\u003e7] for childbirth\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 3 presents the accuracy of the administrative data algorithms in identifying women with childbirth experience and the month and year of their deliveries based on the DeSC database. The accuracy assessment of Algorithm 4, using diagnosis-related codes with a suspected flag or medical procedure codes regarding childbirth, showed the highest value, with a sensitivity of 65.8%, specificity of 99.0%, PPV of 74.4%, NPV of 98.4%, kappa index of 0.69, and Youden Index of 0.65, respectively. Algorithm 6, which was created using diagnosis-related codes or medical procedure codes, resulted in a sensitivity of 63.9%, specificity of 99.0%, PPV of 74.7%, and Kappa Index of 0.68.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidating algorithms [8\u003c/strong\u003e–\u003cstrong\u003e13] for identifying childbirth, excluding potentially preterm birth\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable S6 shows the accuracy of the administrative data algorithms in identifying women with childbirth experience and the month and year of their deliveries excluding those with a high possibility of preterm birth. Algorithm 11, which was created using diagnosis-related codes with a suspected flag or medical procedure code, showed the highest value, with a sensitivity of 65.3%, specificity of 99.0%, PPV of 74.8%, NPV of 98.4%, Kappa Index of 0.68, and Youden Index of 0.64, respectively. Algorithm 13, which was created using diagnosis-related codes or medical procedure codes, resulted in a sensitivity of 63.6%, specificity of 99.0%, PPVs of 75.0%, and Kappa Index of 0.68, respectively.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidating algorithms for identifying second childbirth\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable S7 shows the accuracy of the administrative data algorithms in identifying second childbirth when the evaluation of the algorithms began 2–24 months after the first childbirth during the observation period. When the observation of diagnosis-related codes began 11 months after the first childbirth in Algorithm 4, the Youden Index was 0.5519, the highest value, and the sensitivity, specificity, PPV, NPV, and Kappa Index were 57.7%, 97.5%, 80.8%, 92.6%, and 0.62, respectively.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we established and validated algorithms for defining the month and year of childbirth based on a claims database among women, regardless of their childbirth history. Algorithm 4 involving diagnosis-related codes with a suspected flag or medical procedure codes regarding childbirth yielded the highest value, with a Kappa Index of 0.69 and a sensitivity of 65.8% for all childbirths. The PPV was 74.4% in all algorithms, with the highest value being 77.0%, showing minimal difference. Algorithm 11 excluded births with a high possibility of being preterm from Algorithm 4. Algorithms 4 and 11 showed a nearly similar Youden Index at 0.648 and 0.643, respectively. Considering second childbirth using the claims database, at the start of observation, 11 months after the first childbirth, algorithm 4 yielded the highest Youden Index at 0.551. Although sensitivity was not high at 57.7%, specificity, PPV, and NPV were high at 97.5%, 80.8%, and 92.6%, respectively.\u003c/p\u003e\n\u003cp\u003eAlgorithms 1–7 yielded PPV and NPV of approximately 75% and 95%, respectively, for all childbirths. Algorithm 4, with the highest Kappa Index at 0.69, was the most accurate for identifying childbirth based on claims data. Comparing algorithms 4 and 6, using diagnosis-related codes with a suspected flag yielded a higher sensitivity than without a suspected flag, and both categories had an almost similar specificity. The PPVs of algorithms 4 and 6 were almost the same because there were very few cases of diagnosis-related codes with a suspected flag regarding childbirth among non-pregnant women.\u003c/p\u003e\n\u003cp\u003eMoreover, we established and validated an algorithm for estimating the date of second childbirth. In the analysis, 11 months after the first childbirth, algorithm 4 yielded the highest Youden Index at 0.5519. Women who do not breastfeed after childbirth can ovulate and become pregnant 6 weeks after childbirth [10]. Considering that a baby is born at full term in 37–41 weeks, it takes approximately 11 months to give birth again and at least 6 weeks to become pregnant again, which may explain why the 11-month Youden Index was the highest.\u003c/p\u003e\n\u003cp\u003eIn previous studies, the identification of date of childbirth\u0026nbsp;varied based on factors such as childbirth-related claims data, hospitalisations for childbirth, and birth certificate records, and the validity of algorithm for defining childbirth were based on electronic medical records and birth certificate records\u0026nbsp;[5-8, 11-14].\u0026nbsp;All these studies were conducted only among women who had given birth. In a previous study conducted in Japan, the algorithm was validated using medical claims data of obstetric patients at a single institute. This previous validation study included only women who had given birth and was conducted at a high-level medical institution [15]. Under this condition, it is possible that there was a bias, such as with high-risk pregnancies. All these previous studies were conducted among women who had given birth, and there have been no validation studies on childbirth with established algorithms for defining date of childbirth, regardless of whether they had given birth or not. Therefore, our study targeting women in the general population, with a less-biased claims database, is significant.\u003c/p\u003e\n\u003cp\u003eIn this study, the proportions of mothers aged 15–19, 20–24, 25–29, 30–39, 40–44, and 45–49 years at the time of childbirth were 1.0%, 8.6%, 24.1%, 34.7%, 24.7%, 6.6%, and 0.3%, respectively. The proportions of mothers aged 15–19, 20–24, 25–29, 30–39, 40–44, and 45–49 years at the time of childbirth in Japan in 2022 were 0.6%, 6.9%, 26.3%, 36.3%, 23.8%, 6.0%, and 0.2%, respectively [18]. The proportions reported in our study and Japan were nearly similar; therefore, the inclusion of mothers in this study was thought to be less biased.\u003c/p\u003e\n\u003cp\u003eThe main strength of this study is that it is the first validation study of date of childbirth in a large population of over 850,000 women regardless of birth status. The algorithm based on the claims data in this study would be widely useable and applicable to other claims databases that do not contain the parent–child information that identified childbirth. ICD-10 codes were used in this study; therefore, this algorithm is applicable globally.\u003c/p\u003e\n\u003cp\u003eThis study has several limitations. First, the validation analysis only included women with the parent–child IDs as family information. However, the proportion of women who gave birth was almost the same as the national average; therefore, bias was considered small. Second, women with normal deliveries could not be evaluated by the algorithms in this study because medical insurance is generally not used for normal deliveries. The algorithms used in this study cannot identify childbirths without medical procedures. Therefore, when targeting normal deliveries that do not involve medical procedure codes or diagnosis-related codes regarding childbirth, it is necessary to use administrative databases such as parent–child IDs. Third, there may have been a misallocation of parent–child IDs when linking parent-child relationships in the administrative database. However, the possibility of a link error was considered extremely low because four women aged under 15 (0.02%) and 64 women aged over 50 (0.01%) years were identified as mothers by parent–child IDs.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn this study, we established an optimal algorithm for estimating childbirth based on claims data. This validated algorithm will be useful for accurate estimation of childbirth in future claims database studies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eICD-10, International Classification of Diseases, Tenth Revision; ID, identifier; TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; PPV, positive predictive value; NPV, negative predictive value; 95% CI, 95 percent confidence interval\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of Nara Medical University (approval no. 1123), and conducted in accordance with the principles outlined in the Declaration of Helsinki. All patient data were anonymised prior to analysis. The requirement for informed consent was waived because of the retrospective nature of the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available from DeSC Healthcare, Inc., but restrictions apply to the availability of these data, which were used under the licence for the current study and are not publicly available. However, the data are available from the authors upon reasonable request and with permission from DeSC Healthcare, Inc.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYN received speaker fees from Novo Nordisk, Sanofi, Daiichi Sankyo, and DeSC Healthcare. YT received consultant fees from Novo Nordisk, Otsuka, and Recordati, and speaker fees from Novo Nordisk, Sumitomo Dainippon, Eli Lilly, Ono, Novartis, Nippon Boehringer Ingelheim, AstraZeneca, and Kyowa Kirin. The other authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by a grant from the Japan Ministry of Education, Culture, Sports, Science and Technology. (grant numbers: 22H03355, 23K24613, 24K22341, and 23K16362).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors’ contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eM.K. designed the study, analyzed the data, and wrote the manuscript. H.N. provided advice on this paper and discussed the validity of statistical methods and specific methodologies. Y.N. provided advice on this paper and technical support in statistical methods and extraction of data. Y.T. supervised the study and reviewed the manuscript. E.M., T.N., T.M., and T.I. contributed to writing the manuscript. H.N. and Y.N. contributed equally to this study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe DeSC database was provided by DeSC Healthcare, Inc. Under their academic research support program.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHirose N, Ishimaru M, Morita K, Yasunaga H (2020) A review of studies using the Japanese National Database of Health Insurance Claims and Specific Health Checkups. Annals Clin Epidemiol 2:13\u0026ndash;26\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHall GC, Sauer B, Bourke A, Brown JS, Reynolds MW, LoCasale R (2012) Guidelines for good database selection and use in pharmacoepidemiology research. Pharmacoepidemiol Drug Saf 21:1\u0026ndash;10\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIto F, Togashi S, Sato Y, Masukawa K, Sato K, Nakayama M et al (2023) Validation study on definition of cause of death in Japanese claims data. PLoS ONE 18:e0283209\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFujihara K, Yamada-Harada M, Matsubayashi Y, Kitazawa M, Yamamoto M, Yaguchi Y et al (2021) Accuracy of Japanese claims data in identifying diabetes-related complications. Pharmacoepidemiol Drug Saf 30:594\u0026ndash;601\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRoberts SCM, Schulte A, Zaugg C, Leslie DL, Corr TE, Liu G (2023) Association of pregnancy-specific alcohol policies with infant morbidities and maltreatment. JAMA Netw Open 6:e2327138\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMarchal C, Belhassen M, Guiso N, Jacoud F, Cohen R, Le Pannerer M et al (2022) Cocooning strategy: pertussis vaccination coverage rate of parents with a new-born in 2016 and 2017 in France. Front Pediatr 10:988674\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKim S, Kim C, Kim JH (2024) Antenatal care inequalities in South Korea: an analysis of health insurance claims data (2013\u0026ndash;2022) in a high-resource, high-use country. Int J Gynaecol Obstet 166:718\u0026ndash;726\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSteenland MW, Trivedi AN (2023) Association of Medicaid expansion with postpartum depression treatment in Arkansas. JAMA Health Forum 4:e225603\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKasahara K, Tanaka-Mizuno S, Tsuji S, Ohashi M, Kasahara M, Kawasaki T et al (2024) Pregnancy and lactation-associated osteoporosis as a major type of premenopausal osteoporosis: a retrospective cohort study based on real-world data. BMC Pregnancy Childbirth 24:301\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIshikawa T, Obara T, Jin K, Nishigori H, Miyakoda K, Suzuka M et al (2019) Examination of the prescription of antiepileptic drugs to prenatal and postpartum women in Japan from a health administrative database. Pharmacoepidemiol Drug Saf 28:804\u0026ndash;811\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhu Y, Hampp C, Wang X, Albogami Y, Wei YJJ, Brumback BA et al (2020) Validation of algorithms to estimate gestational age at birth in the Medicaid Analytic eXtract-Quantifying the misclassification of maternal drug exposure during pregnancy. Pharmacoepidemiol Drug Saf 29:1414\u0026ndash;1422\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMoll K, Wong HL, Fingar K, Hobbi S, Sheng M, Burrell TA et al (2021) Validating claims-based algorithms determining pregnancy outcomes and gestational age using a linked claims-electronic medical record database. Drug Saf 44:1151\u0026ndash;1164\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWentzell N, Schink T, Haug U, Ulrich S, Niemeyer M, Mikolajczyk R (2018) Optimizing an algorithm for the identification and classification of pregnancy outcomes in German claims data. Pharmacoepidemiol Drug Saf 27:1005\u0026ndash;1010\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBloti\u0026egrave;re PO, Weill A, Dalichampt M, Billionnet C, Mezzarobba M, Raguideau F et al (2018) Development of an algorithm to identify pregnancy episodes and related outcomes in health care claims databases: an application to antiepileptic drug use in 4.9 million pregnant women in France. Pharmacoepidemiol Drug Saf 27:763\u0026ndash;770\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTajima K, Ishikawa T, Noda A, Matsuzaki F, Morishita K, Inoue R et al (2022) Development and validation of claims-based algorithms to identify pregnancy based on data from a university hospital in Japan. Curr Med Res Opin 38:1651\u0026ndash;1654\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang X, Meisel Z, Kellom K, Whitaker J, Strane D, Chatterjee A et al (2023) Receipt and duration of buprenorphine treatment during pregnancy and postpartum periods in a national privately-insured cohort. Drug Alcohol Depend Rep 9:100206\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePark SJ, Choi NK, Seo KH, Park KH, Woo SJ (2015) Retinal vein occlusion and pregnancy, pre-eclampsia, and eclampsia: the results from a nationwide, population-based study using the national claim database. PLoS ONE 10:e0120067\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVital statistics. Tokyo: Ministry of Health, Labor and welfare (2022) \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mhlw.go.jp/toukei/saikin/hw/jinkou/kakutei22/index.html\u003c/span\u003e\u003cspan address=\"https://www.mhlw.go.jp/toukei/saikin/hw/jinkou/kakutei22/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Accessed 19 Apr 2025\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKundel HL, Polansky M (2003) Measurement of observer agreement. Radiology 228:303\u0026ndash;308\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1. Algorithms for identifying childbirth\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"377\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eCndition A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eDiagnosis-related codes regarding childbirth with a suspected flag\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eCndition B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eDiagnosis-related codes regarding childbirth without a suspected flag\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eCondition C\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eMedical procedure codes regarding childbirth\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eCondition\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eB\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eA and C\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eA or C\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eB and C\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eB or C\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 114px;\"\u003e\n \u003cp\u003eAlgorithm7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 263px;\"\u003e\n \u003cp\u003eC\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. Characteristics of women in the validation study\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"224\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003eNumber\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003ePercentage\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e0～4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e7,162\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.8%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e5～9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e7,201\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.8%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e10～14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e7,505\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.9%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e15～19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e10,278\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e1.2%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e20～24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e20,312\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e2.4%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e25～29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e30,870\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e3.6%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e30～34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e44,518\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e5.2%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e35～39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e53,633\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e6.3%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e40～44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e62,345\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e7.3%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e45～49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e68,259\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e8.0%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e50～54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e65,660\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e7.7%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e55～59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e80,542\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e9.4%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e60～64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e150,979\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e17.7%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e65～69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e193,158\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e22.6%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e70～74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e52,204\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e6.1%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 71px;\"\u003e\n \u003cp\u003e854,626\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e100%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3. Accuracy of administrative data algorithms for identifying women with childbirth experience\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;and the month and year of their deliveries\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm5\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm6\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAlgorithm7\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTP\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e21,854\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e20,947\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e12,279\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e24,957\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e12,083\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e24,223\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e15,123\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e809,177\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e809,613\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e813,011\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e808,126\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e813,073\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e808,500\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e811,960\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e16,080\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e16,987\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e25,655\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e12,977\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e25,851\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e13,711\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e22,811\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFP\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e7,515\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e7,079\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3,681\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e8,566\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e3,619\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e8,192\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e4,732\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity(%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e57.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e55.2%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e32.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e65.8%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e31.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e63.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e39.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e57.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e54.7%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e31.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e65.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e31.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e63.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e39.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e58.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e55.7%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e32.8%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e66.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e32.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e64.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e40.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.5%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.5%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.5%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.2%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e99.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ePPV (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.7%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e76.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e77.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.7%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e76.2%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e73.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.2%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e76.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e76.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e75.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e75.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e77.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e74.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e77.6%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e75.2%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e76.8%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNPV (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e96.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e96.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e96.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e96.9%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.2%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.1%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.0%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e98.4%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e97.3%\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eKappa\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.64\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.62\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.44\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.69\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.44\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.68\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.51\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 115px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eYouden\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.57\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.54\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.32\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.65\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.31\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.63\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 89px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.39\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eAbbreviations: TP, True Positive; TN, True Negative; FP, False Positive; FN, False Negative; 95% CI, 95 percent confidence interval; PPV, positive predictive value; NPV, negative predictive value.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"the Japan Ministry of Education, Culture, Sports, Science and Technology","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Claims data, Childbirth, Validation","lastPublishedDoi":"10.21203/rs.3.rs-8284846/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8284846/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eClaims data analysis is useful in clinical research. However, no validation studies have been conducted using established algorithms to define childbirth among women. The aim of this study was to establish and validate algorithms to define childbirth from a claims database.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThe DeSC database, including claims data for approximately 13\u0026nbsp;million people, as well as parent\u0026ndash;child identifiers (IDs) as family information obtained from insurers, was used. Seven algorithms were designed using combinations of diagnosis-related codes with a suspected flag for childbirth (A), diagnosis-related codes without a suspected flag (B), and medical procedure codes (C). The combinations were A, B, C, A and/or C, and B and/or C. Parent\u0026ndash;child IDs were used to determine the mother\u0026rsquo;s month and year of childbirth based on the child\u0026rsquo;s month and year of birth. The gold standard for the month and year of childbirth was defined as the child\u0026rsquo;s month and year of birth among women aged 15\u0026ndash;49 years linked by parent\u0026ndash;child IDs during the observation period. We calculated sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), Kappa Index, and Youden Index for each algorithm. To validate algorithms for estimating second childbirth during the observation period, which would become useful and valuable in defining childbirth, identification of second childbirth was started 2\u0026ndash;24 months after the first, given that the average age difference was two years.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eA total of 854,626 women were included in this study, of whom 37,934 were aged 15\u0026ndash;49 years at the time of parent\u0026ndash;child ID assignment and classified as experiencing childbirth during the observation period. The algorithm with the highest value was \u0026ldquo;A or C\u0026rdquo; (Kappa Index: 0.69, sensitivity: 65.8%, specificity: 99.0%, PPV: 74.4%, and NPV: 98.4%). For second childbirth, algorithm \u0026ldquo;A or C\u0026rdquo; showed that 11-month difference had the highest Youden Index at 0.551.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eWe developed algorithms based on claims data and established an optimal algorithm for estimating childbirth. This validated algorithm can be used for accurate estimation of childbirth to clarify pregnancy- and childbirth-related diseases in future claims database studies.\u003c/p\u003e","manuscriptTitle":"Appropriate definition of childbirth using Japanese administrative database: A cross-sectional cohort validation study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-11 12:59:40","doi":"10.21203/rs.3.rs-8284846/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3038e052-4127-4a5f-86a4-c4e931f9ced2","owner":[],"postedDate":"December 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":59157213,"name":"Statistical Epidemiology"}],"tags":[],"updatedAt":"2025-12-11T12:59:40+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-11 12:59:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8284846","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8284846","identity":"rs-8284846","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}