Family Vulnerability Scale: evidence of content and internal structure validity

26

territory view based on families’ vulnerability strata allows identifying 27 different health needs that, in their turn, can guide healthcare at primary care scope. 28 Although there are instruments aimed at measuring family vulnerability, they still need 29 robust validity evidences; therefore, they represent a limitation for usability in a country 30 showing multiple socioeconomic and cultural realities, such as Brazil. The present study 31 introduces the development and search for evidences about the validity of the Family 32 Vulnerability Scale for Brazil, known as EVFAM-BR. Methods: items were generated 33 through exploratory qualitative study carried out with 123 professionals. Collected data 34 subsidized the generation of 92 initial items that were subjected to a panel of multi-35 regional and multi-disciplinary judges (n = 73) to calculate the Content Validity Ratio 36 (CVR) – this process resulted in a version of the scale comprising 38 items. 37 Subsequently, it was applied to 1,255 individuals to find evidences about the internal-38 structure validity by using the Exploratory Factor Analysis (EFA). Dimensionality was 39 assessed through Robust Parallel Analysis and the model was tested through cross-40 validation to find EVFAM-BR’s final version. Results: the final version comprised 14 41 items distributed into four domains, with explained variance of 79.02%. All indicators 42 were within adequate and satisfactory limits, without any cross-loading and Heywood 43 Case issues. Reliability indices also reached adequate levels ( α = 0.71; ω = 0.70; glb = 44 0.83 and ORION ranging from 0.80 to 0.93, between domains). Instrument’s score was 45 subjected to normalization, and it pointed towards three vulnerability strata (0 to 4 – 46 Low; 5 to 6 – Moderate; 7 to 14 - High). Conclusion: the scale showed satisfactory 47 validity evidences, which were consistent, reliable ad robust; it led to a synthesized 48 instrument capable of accurately measuring and differentiating family vulnerability in 49 the primary care territory, in Brazil. 50 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 3 51

52 The term vulnerability is the core of debates in different knowledge fields; 53 etymologically it would come from Latin “vulnerare” (harm, impair) and “bile” 54 (susceptible) [1]. However, the understanding about vulnerability shows variations 55 when different dimensions linked to this topic are taken into account [2]. Vulnerability, 56 in the Bioethics field, refers to being in danger or exposed to risk due to individual 57 weakness, which is a feature inherent to human beings [3]. As for the healthcare field, 58 this term has a broader meaning; it is associated with acknowledging humans as 59 susceptible to damage or to risks within the health/illness process due to social 60 disadvantages [4]. 61 Although the literature presents different vulnerability definitions and assesses 62 individual predictors, the concept of family vulnerability is measurable through 63 different ways [4] because it must take into consideration this phenomenon from 64 multiple aspects that, in their turn, are linked to health needs of members from a given 65 family. Aspects related to the health condition of members composing the family 66 nucleus, as well as the community and social context, are elements to be taken into 67 account at the time to investigate vulnerability in families [5,6]. 68 Thus, health services and managers must consider family vulnerability to 69 organize healthcare practices, mainly from the population perspective [7]. Accordingly, 70 territory view from vulnerability strata perspectives subsidizes the process to identity 71 health needs in different population groups [2]. Besides, family vulnerability 72 stratification is essential at the time to plan the offer of services in a given territory, 73 since it would help achieving equity and qualified of population-based care 74 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 4 management. Hence, considering health teams activities and the care provision order in 75 the Health Care Network based on the needs identified through family vulnerability 76 demands, it is needed to deepening in aspects composing family vulnerability. 77 Instruments and initiatives to ensure the observation of aspects likely presented 78 by family vulnerability as components to labor process organizations remain scarce. 79 Some global experiences are linked to this phenomenon: vulnerability-measuring 80

lies on the United Nations Program for Development (UNPD); back in 81 2004, it elaborated the Disaster Risk Index (DRI) to measure and compare countries 82 within a process based on physical, social, economic and environmental factors [8]. 83 Subsequently, in 2006, the Autonomous University of Madrid (Spain) estimated the 84 degree of vulnerability of citizens assumingly susceptible to lack of social protection, 85 based on countries belonging to the Organization for Economic Cooperation and 86 Development (OECD) [9]. With respect to the Latin scene, one research aimed at 87 measuring family vulnerability rates in a Colombian municipality based on a sample of 88 families from all socioeconomic strata living in an urban zone and in a rural one [10]. 89 Other initiatives have been introduced and they aimed at developing instruments 90 to be used by primary healthcare (PHC) teams to measure family vulnerability and, 91 consequently, to contribute to plan healthcare provision in the Brazilian territory [11-92 15]. However, instruments so far developed and used for such a purpose still need 93 robust evidences of their validity, since they present limitations to be used in a country 94 with continental dimensions, and multiple socioeconomic and cultural realities, like 95 Brazil. 96 Thus, it is essential reasoning about the concept of family vulnerability, with 97 emphasis on a diversified population and from this perspective, to develop an 98 instrument which could be used in a standardized way in national scope. Then, the 99 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 5 present study aim to seek validity evidences about the content and internal structure of 100 the Family Vulnerability Scale for Brazil (EVFAM-BR). 101 102

103 The present study followed a psychometric nature design to seek evidences 104 about both content validity (stage 1) and internal structure (stage 2) based on current 105 recommendations by the Educational Research Association (AERA), the American 106 Psychological Association (APA) and the National Council on Measurement in 107 Education (NCME) [16]. The study was approved by the Ethics Research Committee of 108 Hospital Israelita Albert Einstein , which was approved on October 22, 2019 (nº 109 3.674.106, CAAE 12395919.0.0000.0071). 110 111 Stage 1: content validity evidences 112 PHC professionals from all Brazilian geographic regions were invited to join the 113 first qualitative exploratory stage of the study to define the concept of “family 114 vulnerability” and to identify factors likely associated with it, in order to subsidize the 115 development of items for the instrument. It was done to identify different 116 understandings about family vulnerability in different geographic regions countrywide. 117 The invitation was made based on the snowball method [17], by WhatsApp and 118 e-mail. Using the RedCap® electronic tool [18,19], an online semi-structured 119 questionnaire was made available for participants after they read the free consent form 120 and formally accepted to join the study. 121 The questionnaire comprised (i) respondents’ socioeconomic, demographic and 122 health profile identification, (ii) open questions about the concept of vulnerability and 123 scale applicability, and (iii) multiple-choice questions about the relevance of measuring 124 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 6 family vulnerability; the questions were distributed into 46 items elaborated from 125 individual and domestic registration forms used at the Brazilian public healthcare 126 system through e-SUS PHC system [20,21]. Frequencies of responses for items 127 expressed in multi-choice questions and the group of aspects identified in open 128 questions were taken into consideration based on the Content Analysis Technique in 129 order to elaborate the first version of items [22] – they were developed in an 130 interrogative way to allow dichotomous answers (“0 – no” and “1 – yes”). 131 Items developed from the previous stage were subjected to an extensive panel of 132 multi-regional and multi-disciplinary judges. The panel encompassed health 133 professionals, scholars and psychometrists who were invited to join the study through 134 the snowball method. 135 The large number of judges was explained by the need of calculating and 136 applying the instrument at national scope. The judges judged the items in the first 137 version of the instrument based on relevance and clarity criteria, as well as were 138 enquired about the need of changing the writing of any item. Then, option was made to 139 apply the Content Validity Ratio (CVR) scale [23] as the validity index to select the 140 items. CVR is calculated based on the number of judges in the panel [24,25] in order to 141 allow the adoption of a larger number of judges. CVR was initially applied to assess the 142 relevance of a given item in order to check whether it effectively measures the latent 143 variable: family vulnerability. CVR was represented as CVR-1 (the Item’s CVR) and 144 CVR-E (Scale CVR) – this last one corresponds to mean recorded for the CVR criteria. 145 It is important highlighting that a modified CVR version with two points, namely “no” 146 and “yes”, was adopted. The original version had three points and did not have an 147 effective practical effect, since, for CVR calculation purposes, it used to become 148 dichotomous. This procedure was already adopted in studies [26,27]. 149 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 7 Oftentimes, the mean recorded for the assessed questionnaires is adopted; 150 therefore, a hierarchic flow was herein adopted, since other indicators, such as item’s 151 clarity, were only analyzed after judges showed its relevance to testify that the item 152 actually assesses the instrument’s latent variable. The application of requirements in the 153 same stage tends to inflate mean CVR and to launch items that do not measure the latent 154 variable to the next stage. Accordingly, as pointed out by DeVellis [28], a given item 155 can be relevant, but its words might be problematic. Thus, the second stage refers to 156 items’ clarity (whether the item is well written in terms of its semantics). The third stage 157 assessed the need of changing the items’ writing. The mean recorded for CVR was only 158 applied to items that adhered to the phenomenon. 159 160 Stage 2: Evidences about internal structure validity 161 The version subjected to content evidences was applied to users of PHC services 162 to find evidences about internal structure validity. 163 All data collectors were previously trained and clarified about the informed 164 consent form application, as well as about research aim, methodology and questions. 165 Study presentation and data collection flow were also previously carried out with teams 166 from the participating PHC services. 167 PHC services selection was based on municipalities presenting the largest 168 population of attendees of the Program to Support the Institutional Development by the 169 Unified Healthcare System (Proadi-SUS) in Brazil, PlanificaSUS [29]. It was done by 170 including at least one PHC service from each geographic region in the country. Thus, 171 data collection was carried out in 11 PHC services: 1 in Northern Brazil (Roraima 172 State), 1 in the Northeastern region (Pernambuco State) and 2 in Midwestern Brazil 173 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 8 (Mato Grosso State), 5 in the Southeastern region (São Paulo and Minas Gerais states) 174 and 2 in Southern Brazil (Paraná State). 175 The Covid-19 pandemic met the first data collection stage (from June to 176 November 2020); therefore, interviews with São Paulo healthcare unit users were 177 carried out by phone. Service managers were aware of it, since they provided 178 information to identify interviewees in the territory. As for the second data collection 179 time (from May to August 2022), users who had attended the participating PHC service 180 at data-collection day were asked to join the study. 181 Over 18-year-old participants were informed about the informed consent form at 182 both data-collection times and they only joined the research after signing it. 183 Subsequently, the structured questionnaire about the family vulnerability scale was 184 applied, and it was followed by participants featuring in RedCap® [18,19]. 185 186 Statistical analysis 187 Exploratory factor analysis 188 The first stage of the analysis aimed at assessing whether the collected data were 189 prone to factorial through Measure of Sampling Adequacy (MSA). Bartlett sphericity, 190 determinant of the matrix and Kaiser-Meyer-Olkin (KMO) were assessed at this stage. 191 Besides assessing the dataset items, individual analysis was also assessed, as 192 recommended by Lorenzo-Seva and Ferrando [30]. The inadequacy of items to be 193 factored can affect model solution. Missing data were treated through the multiple 194 imputation technique [31]. 195 Dimensionality testing was carried out through Parallel Analysis, based on 196 Optimal implementation of Parallel Analysis (PA) and Minimun rank factor analysis to 197 minimize the common variance of residues [32]. PA was implemented through 198 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 9 permutation with 500 random matrices. Dimensionality in exploratory factorial analysis 199 (unrestricted model) was tested through Parallel Analysis, which has been considered 200 more robust and accurate to test it [33-37]. 201 Tetrachoric matrix estimates were carried out through Bayes Modal Estimation 202 [38], with Smoothing Ridge [39]. The use of tetrachoric/polychoric correlations tends to 203 increase the model’s accuracy in comparison to Pearson’s correlation [40,41]. 204 Factors’ extraction was performed through the RULS technique (Robust 205 Unweighted Least Squares), which reduces the residues in matrices that are more robust 206 in terms of abnormal data [42]. Promin oblique rotation would be used in case the 207 instrument emerged as multi-dimensional [43]. 208 UNICO (Unidimensional Congruence > 0.95), ECV (Explained Common 209 Variance > 0.80 – Quinn, 2014) and MIREAL (Mean of Item Residual Absolute 210 Loadings < 0.30) were adopted as unidimensionality assessment indicator [44]. 211 Quality parameters of the instrument 212 Instrument explained variance must be close to 60% [45]. Initial factorial load of 213 0.30 is recommended when the sample comprises less than 300 individuals [45]; 214 communities must present values higher than 0.40 [46]. The maintenance or removal of 215 a given model item depend on factorial load magnitude, on the communities and on the 216 existence of cross-loading and Heywood cases, as well as on the impermeability of 217 factors. The unique directional correlation (Eta) through Pratt’s Measure was adopted to 218 increase the accuracy of decision-making about the maintenance or removal of a given 219 item [47,48]. 220 Reliability 221 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 10 Reliability was measured through four indicators: Cronbach's alpha [49], 222 Greatest Lower Bound – glb [50], Omega [51] - all three by means of Bayesian 223 estimates - and ORION (Overall Reliability of Fully-Informative prior Oblique N-EAP 224 scores) [52]. 225 Cross-validation was applied to increase the model’s reliability and replicability; 226 the Houdolt technique was also herein applied [53]. This technique divides the dataset 227 into a training sample - that can range from 10%, 30% to 50% - and into a dataset 228 known as test dataset [53]. The dataset in the present study was split in half by 229 randomly choosing the items. The Solomon technique [54] was adopted, so that dataset 230 division could be random and respect factorability’s equivalence. The datasets were 231 labeled as follows: Full Sample (FS; n = 1,255); Training Sample (TrS n = 627) and 232 Test Sample (TsS; n = 628). According to Brown [55], cross-validation can be carried 233 out either through EFA or Confirmatory Factor Analysis (CFA). FS analysis will only 234 take place if the model found in TrS and TsS can be replicated. This procedure was 235 already adopted in previous studies [56,57], and it follows contemporary 236 recommendations [58]. 237 Descriptive study and standardization 238 An exploratory descriptive study of general scores recorded for the Family 239 Vulnerability Scale (FVS - EVFAM-BR, in Brazilian Portuguese) was carried out after 240 a solution for the internal structure was found. Results recorded for the items and for 241 total score were represented by answers’ frequency, median (Md), interquartile interval 242 (IIQ), amplitude (amp), minimum (min) and maximun (max) value. 243 Standardization, in the first stage, was performed by identifying score cuts based 244 on participants’ distribution. Despite this process, although participants’ distribution is 245 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 11 recurrent in standardization studies, it can present distortions, because the score is not 246 directly analyzed, but it can be taken as consequence of participants’ position in the 247 cutting points. Discriminant analysis of each one of the limits and Family Vulnerability 248 Scale scores were used to improve the accuracy of proposed cuts (within the limit) and 249 to assess the predictive ability to classify the individuals. The discriminant analysis aims 250 at better understanding group differences and at predicting the probability of an entity 251 (individual or object) to perceive a specific class or group, based on several independent 252 variables of the metrics [59]. Boedeker and Kearns [60] identified a better performance 253 by the discriminating analysis in comparison to many other techniques applied for the 254 same purpose. Besides, it allows determining the independent variable mostly 255 accounting for differences in mean score profiles in two or more groups [59]. 256 Tabachnick and Fidell [61] added to this information by stating that the aim of the 257 discriminating analysis is to predict the group’s participation based on a set of 258 predictors. Accordingly, it is possible confirming whether the groups formed from the 259 distribution process have properly classified individuals within the established limits. 260 Data were analyzed in statistic software Factor 12.01.01, SPSS v.23 and JASP 261 16.04. 262 263

264 Content validity evidences 265 In total, 123 professionals from the five Brazilian regions joined the first stage of 266 the research to define the concept of “family vulnerability”: 48.8% of them came from 267 Northeastern Brazil; 21.1% from Southeastern Brazil; 17.9% from Southern Brazil, 268 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 12 8.9% from Northern Brazil and 3.3% from Midwestern Brazil. Most professionals 269 belonged to the female sex (82.9%), approximately 40% of them had at least 10-year 270 experience in PHC and 48.8% reported to have specialization degree and higher 271 schooling profiles. Nurses were the professional category accounting for the highest 272 participation in this stage (48.3%); they were followed by Community Health Agents – 273 CHA – (10.7%). The first version of the instrument counted on 92 items; it was 274 developed from factors’ responses that, at first, could be associated with the concept of 275 family vulnerability. 276 A panel of multi-regional and multi-disciplinary judges was set for the second 277 stage; it aimed at identifying content validity evidences. This panel comprised 73 278 judges: 61.7% from Southeastern Brazil, 15.1% from Southern Brazil, 9.6% 279 Northeastern Brazil, 6.8% from Northern Brazil and 6.8% from Midwestern Brazil. 280 Most of them belonged to the female sex (79.5%), more than half of them had less than 281 10-year experience in PHC (57.5%) and specialization as higher schooling profile 282 (51.4%). Nurses were the professional category accounting for the highest participation 283 in the panel (50.7%), they were followed by physicians (16.55). CVR was applied to 284 judges’ answers. CVR critical value was established at CVR > 0.12, which was defined 285 based on the participation of 73 judges. 286 CVR calculation led to the exclusion of 54 items, and it resulted in scale version 287 comprising 38 items linked to socioeconomic and demographic aspects, access to 288 healthcare services, health condition and life style. In order to achieve a better 289 understanding of it, 12 of the 38 items were rewritten based on recommendations from 290 the panel of judges. It must be clear that only items 2 and 15 (Table 1) recorded CVR 291 lower than the critical value; therefore, their text was revised. The other 10 items did not 292 suffer any change; writing adjustments were made in the original text just to meet 293 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 13 CVR’s critical value. The version presenting evidence of content validity (38 items) was 294 taken into consideration in the stage to evidence internal structure validity. 295 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 14 Table 1. Items that remained in the scale after Content Validity Ratio (CVR) application. 296 Item Content Validity Ratio (CVR) Relevance Clarity Need of changeª New text of the item 1. Is there lack of basic sanitation system in the neighborhood you live in? 0.12 0.12 0.78 Is there open sewer in your neighborhood? 2. Do you drink untreated water in your house? 0.12 0.07 0.86 Does the water in your house lack treatment? 3. Does your house face the risk of flood? 0.18 0.40 1.00 - 4. Do you live close to drug dealing areas? 0.12 0.40 0.95 - 5. Does anyone at your house live close to violent people? 0.21 0.32 0.97 - 6. Has anyone in your house been victim of violence? 0.26 0.34 1.00 - 7. Is there violence in your house? 0.23 0.32 1.00 - 8. Is anyone in your house in legal custody condition? 0.12 0.26 0.86 Is anyone in your Family in jail? 9. Is anyone in your house facing financial issues? 0.15 0.12 0.97 - 10. Does anyone lack money to fulfill household needs ? 0.12 0.21 0.97 - 11. Does anyone in your house is a Bolsa Família beneficiary? 0.12 0.32 1.00 - 12. Is anyone in your house a BPC (continued benefit)/LOAS (Social Security Organic Law) beneficiary? 0.15 0.21 0.86 Does anyone in your house get health benefit (BPC /LOAS)? 13. Have any health professional ever mentioned that someone in your house presents obesity? 0.12 0.15 0.86 - . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 15 14. Have any health professional ever mentioned that someone in your house suffers with malnutrition? 0.15 0.12 0.89 - 15. Is there any difficulty in making sure about food variety in your house? 0.15 0.10 0.92 It is hard to make sure about the access to different food types? 16. Does anyone in your house starve? 0.21 0.26 0.92 - 17. Does anyone in your house have drug addiction? 0.23 0.26 0.89 Does anyone in your house use illegal drugs? 18. Is anyone in your house an alcohol abuser? 0.23 0.26 0.92 - 19. Does anyone in your house use controlled medication? 0.15 0.37 0.95 - 20. Does anyone in your house use 5, or more, medications a day? 0.18 0.40 0.95 Does anyone in your house uses 5, or more, medications on a daily basis? 21. Does anyone in your house have a health condition that demands long-term caregiving? 0.21 0.26 0.97 Does anyone in your house have a health condition that requires continuous care? 22. Is anyone in your house impaired to perform daily activities? 0.18 0.23 0.86 - 23. Is anyone in your house helped by others to accomplish its own daily healthcare procedures? 0.15 0.21 0.95 Does anyone in your house need help to accomplish its own daily healthcare procedures? 24. Does anyone in your house present any disability? 0.12 0.37 1.00 - 25. Does anyone in your house have any intellectual/cognitive disability? 0.15 0.29 0.84 - 26. Does anyone in your house have mental issues? 0.21 0.34 1.00 - . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 16 27. Does anyone in your house have HIV/aids? 0.15 0.32 0.97 - 28. Is anyone in your house sick in bed? 0.15 0.32 1.00 - 29. Does anyone in your house often go to urgency and emergency units? 0.18 0.29 0.95 - 30. Does anyone in your house do not know the UBS/healthcare unit team in charge of your family? 0.15 0.32 0.95 - 31. Did anyone in your house have a child without wanting it? 0.15 0.29 0.95 Has anyone in your house had an unplanned child? 32. Did anyone in your house have a child before turning 20 years old? 0.12 0.37 0.97 - 33. Has anyone in your house had its mother absent in childhood? 0.18 0.34 1.00 - 34. Has anyone in your house had an absent father in the childhood? 0.12 0.32 1.00 - 35. Has anyone in your house faced abandonment by the family? 0.21 0.18 0.95 - 36. Are children in your house out of school? 0.12 0.32 0.92 Are there children in your house out of school? 37. Are there adolescents in your hose out of school? 0.12 0.37 0.97 Do you have any adolescent in your house out of school? 38. Are there under 14-year-old individuals in your house that have a job? 0.21 0.23 0.95 - Mean CVR 0.16 0.27 0.94 - ª Accordingly, CVR values higher than the critical value point towards no need of changing the writings, although some items un der this 297 condition were rewritten in order to improve semantics’ adequacy. 298 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 17 299 Evidences about internal structure validity 300 In total, 1,584 users who attended the 11 PHC services during data collection 301 were invited to join this research stage. However, only 1,505 (95%) of them accepted 302 the invitation and signed the informed consent form. Only 1,255 of them completed the 303 interview for the application of the scale version presenting content validity evidences. 304 This sample represented the study’s final sample and this version presented content 305 validity evidences. Table 2 introduces participants’ description of this study stage. 306 Table 2. Participants’ featuring. 307 Variables (n=1255) Categories N (%) Age in yearsa - 43.3 (15.5%) Sex (n=756) Female 551 (43.9%) Male 205 (16.3%) Race/skin color (n=1217) White 386 (30.8%) Brown 640 (51.0%) Black 150 (12.0%) Yellow 23 (1.8%) Indigenous 18 (1.4%) Schooling (in years) (n=1237) 0 to 4 years 160 (12.7%) 5 to 8 years 225 (17.9%) 9 to 11 years 247 (19.7%) 12 to 15 years 480 (38.2%) Over 16 years 125 (10.0%) Job (n=1237) Unemployed or does not have a job 447 (35.6%) Employer 1 (0.1%) Self-employed without social security 111 (8.8%) Self-employed with social security 47 (3.7%) Wage owner 413 (32.9%) retired/pensioner 181 (14.4%) Others 37 (2.9%) income (in minimum wage)b (n=726) No income 58 (4.6%) Up to 1 minimum wage 327 (26.1%) . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 18 > 1 and lower than 2 minimum wages 179 (14.3%) >= 2 and lower than 4 minimum wages 115 (9.2%) >= 4 lower than 10 minimum wages 31 (2.5%) >= 10 minimum wages 4 (0.3%) Does not know 12 (1.0%) Number of children (n=759) 1 child 140 (11.2%) 2 children 198 (15.8%) 3 children 129 (10.3%) More than 3 children 129 (10.3%) Expecting the first child 19 (1.5%) Does not have children 144 (11.5%) Number of households per room in the house (n=1223) 1 120 (9.6%) Private healthcare insurance (n=1231) Yes 180 (14.3%) No 1,051 (83.7%) Systemic High Blood Pressure (n=1240) Yes 212 (16.9%) No 1,028 (81.9%) Diabetes Mellitus (n=1238) Yes 96 (7.6%) No 1,142 (91%) Cancer (current) (n=1237) Yes 6 (0.5%) No 1,231 (98.1%) Heart disease (n=1240) Yes 72 (5.7%) No 1,168 (93.1%) Intellectual/cognitive impairment (n=812) Yes 15 (1.2%) No 797 (63.5%) Tuberculosis (n=1240) Yes 3 (0.2%) No 1,237 (98.6%) Leprosy (n=1239) Yes 1 (0.1%) No 1,238 (98.6%) Kidney issues (n=1237) Yes 35 (2.8%) No 1,202 (95.8%) Breathing issues (n=1232) Yes 96 (7.6%) No 1,136 (90.5%) Mental issues diagnosis (n=1238) Yes 30 (2.4%) No 1,208 (96.3%) a continuous numerical variable described as mean and standard deviation. 308 bone minimum wage corresponds to R$1,212.00 (in Brazilian Real, in 2022). 309 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 19 Factorability 310 The evaluation of sample adequacy measures is the first step of the factorability 311 analysis; it aims at assessing dataset factorability and whether factorial analyses are 312 applicable. Dataset’s general data have shown good factorability: Fs recorded KMO 313 (0.75), Bartlett Sphericity = 6,084.6 (df = 91; P < 0.0001) and determinant of the matrix 314 = 0.00001. As for TrS: KMO (0.74), Bartlett Sphericity = 6,153.7 (df = 91; P < 0.0001) 315 and determinant of the matrix = 0.00001; TsS: KMO (0.71), Bartlett Sphericity = 316 2,617.3 (df = 91; P < 0.0001) and determinant of the matrix = 0.0002. Although general 317 indices presented good indicators, 4 of the 38 initial items have shown factorability 318 issues in three datasets (27 - Does anyone in your house have HIV/aids?; 36 - Are there 319 children in your house out of school?; 37 - Do you have any adolescent in your house 320 out of school?; and 38 - Are there under 14-year-old individuals in your house that have 321 a job?). They were excluded from the analyses based on recommendations by Lorenzo-322 Seva and Ferrando [30]. 323 Dimensionability 324 The first analyses were carried out in TrS. Dimensionality analyzed through 325 parallel analysis pointed towards a 4-dimension model. Closeness of dimensionality 326 values kept the indication for multi-dimensional model: Single = 0.82; ECV = 0.65 and 327 MIREAL = 0.37. Thirteen (13) of the 34 items forming the initial analysis did not 328 present substantial factorial load in the model. Accordingly, the process to remove items 329 in order to adjust the model followed two principles: quantitative (statistical adjustment) 330 and qualitative (interpretability) adjustment. The choice for removing an item was 331 carried out by taking into consideration the set of primary indicators: factorial load, 332 communality, Eta of Pratt’s Importance Measure, existence of cross-loading, Heywood 333 case and model adjustment indices. The items were removed from the scale up to the 334 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 20 time the two principles were congruent to each other, and it resulted in a model for 4-335 dimension TrS, with 14 items with proper statistical adjustment, open for 336 interpretability. The parallel analysis kept on pointing out a 4-dimension solution and 337 explained variance of 78.66%. This model was replicated in TsS and FS. Both datasets 338 confirmed the 4-dimension model; furthermore, the closeness of dimensionality values 339 reinforced the multi-dimensional model (Table 3) for the three datasets. 340 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 21 Table 3. Items’ I-Unico, I-ECV and I-Real values 341 Item I-UNICO I-ECV I-REAL TrS TsS FS TrS TsS FS TrS TsS FS 1. Is anyone in your house facing financial issues? 1.000 0.958 1.000 0.991 0.770 0.984 0.063 0.352 0.084 2. Do you lack money to fulfill household needs? 1.000 0.931 1.000 0.996 0.719 0.990 0.038 0.389 0.061 3. It is hard to make sure about access to different food types? 1.000 0.988 1.000 0.976 0.865 0.992 0.094 0.232 0.053 4. Does anyone in your house use controlled medication? 0.957 0.983 0.970 0.766 0.844 0.799 0.328 0.275 0.320 5. Does anyone in your house use 5, or more medications, on a daily basis? 0.662 0.988 0.926 0.469 0.863 0.711 0.491 0.261 0.361 6. Does anyone in your house have a health condition that requires continuous care? 0.927 0.968 0.963 0.711 0.795 0.782 0.449 0.383 0.374 7. Is anyone in your house impaired to perform daily activities? 0.850 0.966 0.941 0.617 0.789 0.735 0.560 0.400 0.439 8. Does anyone in your house need help to accomplish its own daily healthcare procedures? 0.742 0.961 0.887 0.525 0.777 0.657 0.538 0.378 0.478 9. Did anyone in your house have its mother absent in childhood? 0.291 0.347 0.387 0.233 0.270 0.295 0.600 0.493 0.516 10. Did anyone in your house have an absent father in childhood? 0.495 0.240 0.511 0.363 0.198 0.373 0.465 0.535 0.389 11. Has anyone in your house faced abandonment by the family? 0.927 0.583 0.912 0.711 0.418 0.690 0.329 0.439 0.297 12. Does anyone in your house live close to violent people? 0.868 0.936 0.488 0.636 0.727 0.358 0.481 0.115 0.616 13. Have anyone in your house been victim of violence? 0.861 0.914 0.731 0.628 0.693 0.517 0.450 0.295 0.564 14. Is there violence in your house? 0.914 0.963 0.349 0.692 0.782 0.272 0.381 0.021 0.503 I-UNICO, Unidimensional Congruence; I-ECV, Explained Common Variance; I-REAL, Residual Absolute Loadings; TrS, Training Sample; TsS, Test Sample; FS, Full Sample. . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 22 342 TrS primary data (Table 4) presented factorial data ranging from 0.597 to 0.975, 343 communality ranging from 0.449 to 0.967, and Eta ranging from 0.640 to 0.941. The 344 model recorded explained variable of 76.18%. Items 1 and 3 comprised dimension 345 Income, items from 4 to 8 comprised dimension Healthcare, dimension Family was in 346 items 9 to 11, and the single dimension called Violence was observed in items 12 to 14. 347 Accordingly, the model points towards good factorial and interpretable (quantitatively) 348 solution, with content alignment in coherent and interpretable (quantitatively) items. 349 The final version of the scale represented reduction by approximately 63% in the 38 350 items assessed through judges’ panel in the first stage. This value is close to that 351 presented by DeVellis [28], according to whom the researcher must project items lost by 352 50% throughout the process. 353 Based on the four dimensions composing the scale, it is possible taking into 354 account the key role played by social determinants within the health/illness process. 355 Dimensions embody items related to income, social and family cohesion, and to life and 356 housing conditions associated with psychosocial and behavioral aspects [62]. Thus, it 357 extrapolates the biological view of health, which is overall acknowledged in a 358 reductionist way, centered in medical practices [62-65]. Nevertheless, the present 359 instrument emerges as multi-disciplinary work-tool available for PHC teams that have 360 the potential to promote social justice by taking into consideration social inequities at 361 the time to plan healthcare services. 362 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 23 Table 4. Training dataset: Factorial loads, communality and Eta. 363 Item Factorial Load h2 Pratt’s Measure - (Eta) Income Healthcare Family Violence Income Healthcare Family Violence 1. Is anyone in your house facing financial issues? 0.893 0.027 -0.012 0.059 0.849 0.904 0.098 0.000 0.146 2. Do you lack money to fulfill household needs? 0.895 -0.013 0.013 -0.067 0.762 0.872 0.000 0.049 0.000 3. Is there any difficulty in making sure about food variety in your house? 0.597 0.091 0.102 0.110 0.516 0.642 0.180 0.174 0.203 4. Does anyone in your house use controlled medication? -0.100 0.693 0.101 0.067 0.500 0.000 0.680 0.146 0.130 5. Does anyone in your house use 5, or more, medications on a daily basis? -0.080 0.711 0.031 -0.118 0.449 0.000 0.668 0.051 0.000 6. Does anyone in your house have a health condition that requires continuous care? -0.032 0.805 0.020 0.069 0.671 0.000 0.805 0.059 0.139 7. Is anyone in your house impaired to perform daily activities? 0.094 0.862 -0.131 0.029 0.798 0.189 0.869 0.000 0.084 8. Does anyone in your house need help to accomplish its own daily healthcare procedures? 0.062 0.737 -0.053 -0.090 0.540 0.132 0.723 0.000 0.000 9. Did anyone in your house have its mother absent in childhood? 0.012 -0.140 0.798 0.013 0.630 0.042 0.000 0.790 0.061 10. Did anyone in your house have its father absent in childhood? 0.056 -0.033 0.741 -0.097 0.514 0.099 0.000 0.710 0.000 11. Has anyone in your house faced abandonment by the family? 0.018 0.148 0.630 0.073 0.504 0.065 0.203 0.658 0.163 12. Does anyone in your house live close to violent people? 0.192 -0.130 -0.108 0.975 0.967 0.284 0.000 0.000 0.941 13. Have anyone in your house been victim of violence? -0.118 0.086 0.290 0.605 0.574 0.000 0.146 0.379 0.640 14. Is there violence in your house? -0.124 0.055 -0.091 0.937 0.787 0.000 0.117 0.000 0.880 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 24 364 Factorial loads recorded for the TsS dataset (Table 5) ranged from 0.643 to 365 0.976, communality ranged from 0.482 to 0.910 and Eta ranged from 0.658 to 0.951. 366 This model presented explained variance of 76.18%. Once again, the observed model 367 was equal to the training dataset model; consequently, it was quantitatively and 368 qualitatively interpretable. 369 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 25 Table 5. Test dataset: Factor loading, communality and Eta 370 Item Factorial Load h2 Pratt’s Measure - (Eta) Income Healthcare Family Violence Income Healthcare Family Violence 1. Is anyone in your house facing financial issues? 0.801 0.032 0.046 0.056 0.702 0.818 0.108 0.128 0.075 2. Do you lack money to fulfill household needs? 0.976 -0.072 0.013 -0.008 0.910 0.951 0.000 0.068 0.000 3. Is there any difficulty in making sure about food variety in your house? 0.805 0.063 -0.084 -0.006 0.647 0.790 0.151 0.000 0.000 4. Does anyone in your house use controlled medication? 0.059 0.671 -0.009 -0.006 0.482 0.138 0.681 0.000 0.000 5. Does anyone in your house use 5, or more, medications on a daily basis? 0.141 0.639 -0.078 0.066 0.493 0.227 0.658 0.000 0.094 6. Does anyone in your house have a health condition that requires continuous care? -0.047 0.882 0.023 -0.012 0.753 0.000 0.865 0.063 0.000 7. Is anyone in your house impaired to perform daily activities? 0.001 0.881 -0.030 0.011 0.770 0.015 0.876 0.000 0.035 8. Does anyone in your house need help to accomplish its own daily healthcare procedures? -0.080 0.856 0.037 -0.033 0.690 0.000 0.827 0.078 0.000 9. Did anyone in your house have its mother absent in childhood? -0.024 -0.030 0.727 0.036 0.514 0.000 0.000 0.715 0.059 10. Did anyone in your house have its father absent in childhood? -0.015 -0.090 0.752 0.037 0.546 0.000 0.000 0.736 0.059 11. Has anyone in your house faced abandonment by the family? -0.035 0.077 0.757 -0.062 0.575 0.000 0.123 0.748 0.000 12. Does anyone in your house live close to violent people? 0.067 -0.016 -0.037 0.918 0.843 0.076 0.000 0.000 0.915 13. Have anyone in your house been victim of violence? 0.138 0.096 0.310 0.643 0.652 0.211 0.164 0.370 0.666 14. Is there violence in your house? -0.177 -0.041 -0.172 0.923 0.893 0.196 0.042 0.167 0.909 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 26 371 These results ranged from 0.308 to 0.785 for factorial loads, from 0.212 to 0.967 372 for communality, and Eta ranged from 0.640 to 0.941. The model based on the total 373 sample recorded explained variance of 79.02%. Number of dimensions’ stability and 374 model interpretability are essential aspects of dimensionality. It reinforces the relevance 375 of carrying out the cross-validation. 376 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 27 Table 6. Full dataset: Factor loading, communality and Eta. 377 Item Factorial load h2 Pratt´s Measure - (Eta) Income Healthcare Family Violence Income Healthcare Family Violence 1. Is anyone in your house facing financial issues? 0.733 0.009 0.004 0.023 0.548 0.736 0.049 0.029 0.047 2. Do you lack money to fulfill household needs? 0.785 -0.045 -0.015 -0.020 0.585 0.765 0.000 0.000 0.000 3. Is there any difficulty in making sure about food variety in your house? 0.555 0.059 0.019 0.027 0.346 0.569 0.124 0.060 0.050 4. Does anyone in your house use controlled medication? -0.001 0.523 0.056 0.016 0.287 0.000 0.528 0.087 0.035 5. Does anyone in your house use 5, or more, medications a day? 0.019 0.489 -0.012 -0.010 0.243 0.060 0.489 0.000 0.000 6. Does anyone in your house have a health condition that requires continuous care? 0.007 0.639 0.032 0.013 0.421 0.041 0.644 0.065 0.032 7. Is anyone in your house impaired to perform daily activities? 0.008 0.723 -0.048 0.001 0.518 0.044 0.718 0.000 0.008 8. Does anyone in your house need help to accomplish its own daily healthcare procedures? -0.036 0.624 -0.017 -0.024 0.371 0.000 0.609 0.000 0.000 9. Did anyone in your house have its mother absent in childhood? -0.027 -0.049 0.594 0.004 0.339 0.000 0.000 0.582 0.013 10. Did anyone in your house have an absent father in childhood? 0.009 -0.043 0.552 -0.009 0.302 0.037 0.000 0.548 0.000 11. Has anyone in your house faced abandonment by the family? 0.004 0.060 0.471 0.001 0.236 0.026 0.089 0.476 0.007 12. Does anyone in your house live close to violent people? 0.025 -0.034 0.011 0.743 0.554 0.047 0.000 0.029 0.742 13. Has anyone in your house been victim of violence? 0.028 0.084 0.273 0.308 0.212 0.069 0.119 0.296 0.325 14. Is there violence in your house? -0.047 -0.005 -0.018 0.597 0.351 0.000 0.000 0.000 0.593 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 28 378 Reliability indices between analysis datasets ranged from 0.69 to 0.71 in 379 Cronbach’s alpha, it reached 0.70 in the three dataset for Omega, it ranged from 0.83 to 380 0.84 for glb and from 0.80 to 0.96 ORION, between dimensions and datasets. Factorial 381 solution quality indices also showed adequate levels, and this finding reinforced the 382 model’s stability (Table 7). Accordingly, the set of applied techniques and indices 383 pointed towards a set of internal structure validity evidences that are adequate, 384 consistent, robust and interpretable. 385 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 29 Table 7. Synthesis of the models 386 Synthesis Index Technique Training Sample (TrS) Test Sample (TsS) Full Sample (FS) Exploratory Adequacy of correlation matrix Determinant of the matrix < 0.000001 < 0.000001 0.0002 Bartlett 6084.6 (df = 91) 6153.7 (df = 91) 8936.7 KMO (Kaiser-Meyer-Olkin) 0.75 0.74 0.71 Explained Variance (AP) 77.73% 76.18% 79.02% Polychoric Correlation (rp = ) -0.08 to 0.80 -0.21 to 0.81 -0.04 to 0.82 Reliability Cronbach's Alpha 0.69 0.71 0.71 McDonald's Omega 0.70 0.70 0.70 Greatest Lower Bound – glb 0.83 0.84 0.83 ORIONª 0.90; 0.96; 0.80; 0.91 0.81; 0.91; 0.93; 0.93 0.80; 0.93; 0.90; 0.92 Unidimensional Assessment Unidimensional Congruence (UNICO) 0.82 0.83 0.79 Explained Common Variance (ECV) 0.65 0.70 0.66 Mean of item residual absolute loading (MIREAL) 0.37 0.32 0.36 Quality and Effectiveness Factor Determinacy Index (FDI)ª 0.94; 0.98; 0.89; 0.95 0.90; 0.95; 0.96; 0.96 0.89; 0.96; 0.95; 0.95 Sensivity Ratio (SR)ª 3.02; 5.37; 2.04; 3.20 2.09; 3.29; 3.77; 3.90 2.00; 3.74; 3.11; 3.37 Expected percentage of true differences (EPTD)ª 92.6%; 96.6%; 88.9%; 93.0% 89.1%; 93.3%; 94.3%; 94.6% 88.7%; 94.2%; 92.8%; 93.5% ª from dimension 1 to 4, respectively . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 30 Standardization 387 The internal structure of the instrument was extrapolated and found. Now, the 388 descriptive study and score standardization will be addressed to allow instrument 389 interpretability and participants’ proper classification based on the scores. Accordingly, 390 Table 8 depicts the frequency of answers to items in the questionnaire. There was clear 391 prevalence of “No” answers for all items in the instrument. Some items presented 392 higher frequency of “yes” answers: “Lack of money to fulfill household needs” 393 (40.34%), “someone in the house uses controlled medication” (36.41%), “someone in 394 the house has a health condition that requires continuous care” (36.77%) and “someone 395 in the house had an absent father in childhood” (34.07%). 396 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 31 Table 8. Frequency of answers for the items. 397 Item Frequency of answer to the item (N/%) No Yes Missing 1. Is anyone in your house facing financial issues? 870 (69.10) 377 (29.94) 12 (0.95) 2. Do you lack money to fulfill household needs? 742 (58.93) 508 (40.34) 9 (0.71) 3. Is there any difficulty in making sure about food variety in your house? 928 (73.70) 322 (25.57) 9 (0.71) 4. Does anyone in your house use controlled medication? 793 (62.98) 461 (36.61) 5 (0.39) 5. Does anyone in your house use 5, or more, medications a day? 1,020 (81.01) 231 (18.34) 8 (0.63) 6. Does anyone in your house have a health condition that requires continuous care? 788 (62.58) 463 (36.77) 8 (0.63) 7. Is anyone in your house impaired to perform daily activities? 1,026 (81.43) 232 (18.42) 1 (0.07) 8. Does anyone in your house need help to accomplish its own daily healthcare procedures? 1,066 (84.67) 190 (15.09) 3 (0.23) 9. Did anyone in your house have its mother absent in childhood? 1,048 (83.24) 207 (16.44) 4 (0.31) 10. Did anyone in your house have an absent father in childhood? 826 (65.60) 429 (34.07) 4 (0.31) 11. Has anyone in your house faced abandonment by the family? 1,129 (89.67) 125 (9.92) 5 (0.39) 12. Does anyone in your house live close to violent people? 1,221 (96.98) 34 (2.70) 4 (0.31) 13. Has anyone in your house been victim of violence? 1,088 (86.41) 167 (13.26) 4 (0.31) 14. Is there violence in your house? 1,232 (97.85) 25 (1.98) 2 (0.15) . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 32 398 Table 9 presents the scores recorded for the dimensions and the general score of 399 the Family Vulnerability Scale. All dimensions had all their amplitudes answered. 400 Dimensions Income, Family and Violence recorded median = 0, Healthcare showed 401 median = 1 and total score recorded median = 2. An interesting aspect of the total score 402 lies on the fact that amplitude ranged from 0 to 14 and the maximum score recorded in 403 the current sample reached 12. Medians in the minimum limit, and close to it, 404 previously pointed out that the instrument can accurately differentiate individuals who 405 are eventually facing family vulnerability situations. 406 Table 9. Description of dimensions and scores recorded for the Family 407 Vulnerability Scale. 408 Dimension / Score Central Trend Measurements and Dispersion Median Minimum Maximum Amplitude Interquartile Income dimension 0.00 0.00 3 3 2.00 Healthcare dimension 1.00 0.00 5 5 2.00 Family dimension 0.00 0.00 3 3 1.00 Violence Dimension 0.00 0.00 3 3 0.00 Total Score Total 2.00 0.00 12 12 3.00 409 Because these scores are closer to the minimum limit, they only started 410 presenting greater difference when they got far from the median that, in this case, was 411 close to the minimum; therefore, in the upper quartile. Thus, three initial classifications 412 were suggested: model 1 had cut in the median (low and high vulnerability), model 2 413 scores were separated until percentile 75, from 76 to 89, and higher than 90; model 3 414 scores were separated up to percentile 70, from 70 to 89, or higher. 415 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 33 The discriminant analysis of each classification developed to assess whether it 416 was possible accurately identifying participants within the limit was applied after the 417 first cuts were made. 418 The first analysis adopted a binary classification (low and high). The 419 discriminant analysis showed MBox = 446.58 p < 0.001. λ wilks = 0.36; F (1, 1257) = 420 1,268.77; p < 0.001; canonical correlation = 0.797; the model with two limits properly 421 classified 85.7% of the cases. The discriminant analysis applied to model 2 was MBox 422 = 49.64 p < 0.001. λ wilks = 0.18; F (2, 1256) = 2,094.25; p < 0.001; canonical correlation = 423 0.907. Model 2 properly classified 100% of cases. The analysis applied to model 3 424 presented MBox = 49.64 p < 0.001. λ wilks = 0.25; F (2, 1256) = 1,838.71; p < 0.001; 425 canonical correlation = 0.863; it was possible properly classifying 89% of cases. The 426 recommended classification and scores interpretations are depicted in Table 10. 427 Table 10. Limits, classification and interpretation of Family Vulnerability Scale 428 scores 429 Classification results Percentile Name Score Limit Up to 75 low 0 to 4 76 to 89 Moderate 5 to 6 Higher than 90 High Higher than 7 430 Family Vulnerability Scale – final version 431 The final version of the Family Vulnerability scale (EVFAM-BR) comprised 14 432 items (or questions) applied to each family in the PHC territory in Brazil, due to the 433 action by Community Health Agents (CHA). EVFAM-BR has four dimensions; each 434 one of them has a score corresponding to the number of items in the dimension – at the 435 end of its application, the score must range from zero (0) to fourteen (14). Healthcare is 436 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 34 the dimension accounting for the largest number of items; consequently, it has the 437 greatest potential in the scale (n=5). The final EVFAM-BR presented three family 438 vulnerability classification limits if one sums the scores of each dimension: Low (0 to 439 4), Moderate (5 to 6) and High (7 to 14). Table 11 presents a summary of the Family 440 Vulnerability Scale, and its respective dimensions, items and scores. 441 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 35 Table 11. Family Vulnerability Scale (EVFAM-BR). 442 Dimension Item Item score Dimension score Income 1. Is anyone in your house facing financial issues? 1 3 2. Do you lack money to fulfill household needs? 1 3. Is there any difficulty in making sure about food variety in your house? 1 Healthcare 4. Does anyone in your house use controlled medication? 1 5 5. Does anyone in your house use 5, or more, medications on a daily basis? 1 6. Does anyone in your house have a health condition that requires continuous care? 1 7. Is anyone in your house impaired to perform daily activities? 1 8. Does anyone in your house need help to accomplish its own daily healthcare procedures? 1 Family 9. Did anyone in your house have its mother absent in childhood? 1 3 10. Did anyone in your house have an absent father in childhood? 1 11. Has anyone in your house faced abandonment by the family? 1 Violence 12. Does anyone in your house live close to violent people? 1 3 13. Has anyone in your house been victim of violence? 1 14. Is there violence in your house? 1 Total 14 14 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 36 443

444 The Family Vulnerability Scale (EVFAM-BR) has shown evidences of content 445 and internal structure validity based on the multi-regional and multi-professional 446 context, and this finding allows measuring family vulnerability in Brazil. 447 EVFAM-BR comprises 14 items distributed into the following dimensions: 448 Income, Healthcare, Family and Violence. It is worth highlighting that EVFAM-BR 449 aims at measuring social vulnerability within the family context; consequently, all items 450 refer to the family nucleus, they are not oriented to one specific resident, or to the 451 respondent, itself. 452 Just to exemplify EVFAM-BR application to a family in a given local or time: 453 one of the residents in a given house is facing financial issues. However, lack of money 454 to fulfill household needs is not identified and there is no hard time accessing different 455 food types; one of the residents has a chronic disease that requires continuous care and 456 uses controlled medication (less than 5 medication types a day), but none of the 457 residents has any difficulty in performing daily activities and does not need daily 458 healthcare; none of the residents lacked mother or father presence in childhood and no 459 family member faced abandonment situations; there was no violence in the house and 460 no one in the house lives with violent people, but one of the residents was a victim of 461 violence. Given the positive answers to items financial issue by one of the residents (1), 462 health condition requiring continuous care (1), the use of medication (1) and person 463 who was violence victim (1), this family would reach score 4, and – based on the 464 classification limit of EVFAM-BR final score - it represents a family classified as “low 465 family vulnerability”. 466 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 37 The literature consistently states that income is a relevant social health 467 determinant and that it must be taken into account to allow planning equitable 468 healthcare provision [66]. The subsequent discussion about this topic led to different 469 views on how income inequality affects health. Rich or poor individuals, in low social 470 cohesion societies, would be the target of problems such as crime, lack of public 471 investments, and it makes people adopt unhealthy behaviors such as smoking, excessive 472 alcohol consumption and sedentary life [67]. These outcomes can help better 473 understanding the relationship between variables linked to income and enable 474 interventions at macroeconomic level, as well as assessing these changes in population 475 health. 476 The healthcare dimension is timely, given the accelerated population aging in 477 the country and abroad, a fact that demands healthcare services’ reorganization to 478 continuously fulfill population needs, in an organized way, based on quality and safety. 479 With respect to the family dimension, several studies have shown the association 480 between absence of parents and different health outcomes, among them one finds 481 cognitive development loss [68,69], impacts on mental health [70,71], and early 482 development of risk behavior for health, such as smoking and alcohol abuse [72]. 483 Finally, dimension “violence” corroborated the discussion observed in WHO’s 484 2030 agenda for the sustainable development of millennium goals; this agenda 485 highlights violence prevention as fundamental component for both development and 486 improved quality of life, worldwide. It is known that violence affects health and 487 broadens the demands for healthcare in a way wider than simply through initial trauma, 488 it extrapolates the probability of other important causes for diseases and death [73]. It is 489 possible identifying association among exposure to violence, undesired health outcomes 490 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 38 and unhealthy behaviors, such as drug and alcohol abuse, mainly among low-income 491 mothers in urban locations [74]. 492 Although the urgency in having research focused on low and medium income countries 493 that account for 90% of the global violence, only 10% of studies in this field are 494 performed in them [73]. Thus, EVFAM-BR emerges as a tool to allow structurally and 495 routinely introducing the approach of social factors associated with the health/illness 496 process in healthcare services, in developing countries. 497 Accordingly, EVFARM-BR validity evidences, along with the four family 498 vulnerability strata proposed based on its application, present the potential of this tool to 499 help the role played by PHCs in performing their attributes, mainly in coordinating 500 caregiving based on population-base management within the community and family 501 context [75]. 502 It is important highlighting that EVFAM-BR is an instrument presenting robust 503 and synthesized evidences that, at first, demand low workload investment by 504 professionals and low financial resources. Because it is an objective instrument (only 505 “yes” and “no” answers are expected), it suggests that all PHC professionals must be 506 trained to use it. This instrument emerges as powerful tool to support the work by 507 community health agents (CHA), since these actors are community members and are 508 closely bond to families in the territory; this process makes the “interview environment” 509 more comfortable and trustful for users who answer the questions on behalf of the 510 household. Thus, EVFAM-BR can be applied through printed materials or, yet, online, 511 since it is a tool used as work instrument added to the teams’ routine. One must take 512 into account its potential for inclusion in the digital registration system of the Brazilian 513 Unified Health System (SUS). 514 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 39 It is possible guiding the teams at the time to plan their actions and health 515 interventions (based on exposing families to conditions that increase vulnerability and 516 risk to develop illnesses) by applying EVFAM-BR and by interpreting its household-517 classification results based on the four predicted strata. It is worth highlighting how 518 vulnerability contexts can be changed overtime; instrument application must be 519 periodical to keep teams’ planning updated, according to population needs. 520 Among limitations of the current study, one finds sampling based on 521 convenience. It does not ensure statistical results’ reliability. However, the study was 522 carried out in different socioeconomic, demographic and cultural contexts, since it 523 encompassed participants from the five geographic regions in Brazil. Yet, it is important 524 pointing out the potential of carrying out research in the PHC context in order to allow 525 the participation of people with different demands, needs and life conditions, who seek 526 care in this service. 527 Highlights in the present research are data collection by professionals outside the 528 assessed services who were trained to carry out the interviews, as well as the use of 529 robust techniques to identify EVFAM-BR validity evidences; among them, CVR 530 presents a sophisticated and more adequate method [76] in comparison to the proposed 531 alternatives [77]. CVR calculation takes into consideration the number of judges [24,25] 532 and it minimizes the increase in random compliance [24], a fact that allows adopting a 533 large number of judges to judge the instrument. The adoption of a multi-disciplinary 534 panel of judges comprising researchers, translators, health professionals, methodology 535 experts and lay people leads to more consistent results from the judges’ panel [77-79]. 536 It is important pinpointing the need of implementing research in order to identify 537 the potential and challenges of using EVFAM-BR in the routine of PCH’s services in 538 different Brazilian contexts and, yet, in countries presenting similar health system 539 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 40 features, as well as population socioeconomic, demographic, sanitary and 540 epidemiological features. 541 542

543 Given the set of herein employed techniques, it is possible stating that the set of 544 content validity and EVFAM-BR internal structure evidences are adequate, consistent, 545 reliable and robust, as well as that the cross-validation method ensured model reliability 546 and replicability. A synthetic scale was presented, and it is capable of accurately 547 measuring and differentiating familiar vulnerability. 548 549

550 1. Carmo ME, Guizardi FL. O conceito de vulnerabilidade e seus sentidos para as 551 políticas públicas de saúde e assistência social. Cad. Saude Publica [Internet]. 552 2018 [cited 2022 Nov 22];34(3):e00101417. DOI https://doi.org/10.1590/0102-553 311X00101417. Available at: 554 https://www.scielo.br/j/csp/a/ywYD8gCqRGg6RrNmsYn8WHv/abstract/?lang=555 pt# 556 2. Ferreira JBB, Santos LL, Ribeiro LC, Fracon BRR, Wong S. Vulnerability and 557 Primary Health Care: An Integrative Literature Review. J Prim Care Community 558 Health [Internet]. 2021 [cited 2022 Dec 21];12 DOI 559 10.1177/21501327211049705. Available from: 560 https://pubmed.ncbi.nlm.nih.gov/34654333/ 561 3. Oviedo RAM, Czeresnia D. conceito de vulnerabilidade e seu caráter biossocial. 562 Interface [Internet]. 2015 [cited 2022 Nov 22];19(53):237-250. DOI 563 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 41 https://doi.org/10.1590/1807-57622014.0436. Available at: 564 https://www.scielo.br/j/icse/a/5BDdb5z4hWMNn58drsSzktF/abstract/?lang=pt# 565 4. Ayres JRCM, Calazans GJ, Saletti Filho HC, França Junior I. Risco, 566 vulnerabilidade e práticas de prevenção e promoção da saúde. In: Campos GWS, 567 Bonfim JRA, Minayo MCS, Akerman M, Drumond Junior M, Carvalho YM 568 (eds) orgs. Tratado de Saúde Coletiva. Hucitec Editora; 2016;399-442. 569 5. Ayres JRCM, França Junior I, Calazans GJ, Saletti Filho HC. O conceito de 570 vulnerabilidade e as práticas de saúde: novas perspectivas e desafios. In: 571 Czeresnia D, Freitas CM, editors. Promoção da saúde: conceitos, reflexões, 572 tendências. Rio de Janeiro: Fiocruz; 2003.; p. 117-139. 573 6. Ayres JRCM, Paiva V, França Junior I, Gravato N, Lacerda R, Negra MD, et al. 574 Vulnerability, human rights, and comprehensive health care needs of young 575 people living with HIV/AIDS. Am J Public Health [Internet]. 2006 [cited 2022 576 Nov 14];96(6):1001-1006. DOI 10.2105/AJPH.2004.060905. Available at: 577 https://pubmed.ncbi.nlm.nih.gov/16449593/ 578 7. Silva TMR, Alvarenga MRM, Oliveira MAC. Evaluation of the vulnerability of 579 families assisted in primary care in Brazil. Rev Lat Am Enfermagem [Internet]. 580 2012 [cited 2022 Dec 21];20(5):935-43. DOI 10.1590/s0104-581 11692012000500016. Available from: 582 https://pubmed.ncbi.nlm.nih.gov/23174839/ 583 8. Programa de las Naciones Unidas para el Desarrollo (PNUD), editor. La 584 reducción de riesgos de desastres: un desafío para el desarrollo [bibliography on 585 the Internet]. New York: Programa de las Naciones Unidas para el Desarrollo; 586 2004 [cited 2022 Nov 22]. Available at: 587 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 42 https://unici.gnomio.com/pluginfile.php/18234/mod_folder/content/0/3.25_undp588 _rdr_esp.pdf?forcedownload=1 589 9. Ayuso IA, Vargas EC. Índice de Vulnerabilidad Social en los países de la 590 OCDE. Quivera [Internet]. 2006 [cited 2022 Nov 22];8(2):248-274. Available 591 at: https://www.redalyc.org/pdf/401/40180212.pdf 592 10. Hidalgo-Troya A, Guerrero-Díaz GF, Estupiñan-Ferrín VL, Rocha-Buelvas A. 593 Índice de vulnerabilidad de los hogares en el municipio de Pasto, Colombia, 594 2012. Cad Saude Publica [Internet]. 2017 [cited 2022 Nov 14];33(3):e00122315. 595 DOI 10.1590/0102-311X00122315. Available at: 596 https://www.scielo.br/j/csp/a/FP5Mfs5srgx6CMKX7hf7GGd/?lang=es 597 11. Savassi LCM, Lage JL, Coelho FLG. Sistematização de um instrumento de 598 estratificação de risco familiar: Escala de risco familiar de Coelho-Savassi. J 599 Manag Prim Health Care [Internet]. 2012 [cited 2022 Nov 14];3(2):179-185. 600 DOI https://doi.org/10.14295/jmphc.v3i2.155. Available at: 601 https://www.jmphc.com.br/jmphc/article/view/155 602 12. Amendola F, Alvarenga MRM, Gaspar JC, Yamashita CH, Oliveira MAC. 603 Validade aparente de um índice de vulnerabilidade das famílias a incapacidade e 604 dependência. Rev Esc Enferm USP [Internet]. 2011 [cited 2022 Nov 605 14];45(spe2):1736-1742. DOI https://doi.org/10.1590/S0080-606 62342011000800017. Available at: 607 https://www.scielo.br/j/reeusp/a/WN3Xh9ySynhWsQmKr6zbQVH/abstract/?lan608 g=pt 609 13. Amendola F, Alvarenga MRM, Latorre MRDO, Oliveira MAC. Development 610 and validation of the Family Vulnerability Index to Disability and Dependence 611 (FVI-DD). Rev Esc Enferm USP [Internet]. 2014 [cited 2022 Nov 14];48(1):82-612 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 43 90. DOI 10.1590/S0080-623420140000100010. Available at: 613 https://www.scielo.br/j/reeusp/a/GHNyt8QQbG7FxhFNBWzthpB/?format=pdf614 &lang=en 615 14. Amendola F, Alvarenga MRM, Latorre MRDO, Oliveira MAC. Índice de 616 vulnerabilidade a incapacidades e dependência (IVF-ID), segundo condições 617 sociais e de saúde. Cien Saude Colet [Internet]. 2017 [cited 2022 Nov 618 14];22(6):2063-2071. DOI 10.1590/1413-81232017226.03432016. Available at: 619 https://www.scielo.br/j/csc/a/jtCXRphGrsGjLSTpMsqzZjC/?lang=pt 620 15. Associação Brasileira de Empresas de Pesquisa - ABEP [Internet]. [place 621 unknown]; 2018. Critério Brasil 2018; [cited 2022 Nov 14]; Available at: 622 https://www.abep.org/criterio-brasil 623 16. American Psychological Association - APA [Internet]. Washington: American 624 Educational Research Association (AERA), American Psychological 625 Association (APA), National Council on Measurement in Education (NCME); 626 2014. The Standards for Educational and Psychological Testing; [cited 2022 627 Nov 14]; Available at: https://www.apa.org/science/programs/testing/standards 628 17. Biernacki P, Waldorf D. Snowball Sampling—Problems and Techniques of 629 Chain Referral Sampling. Sociol Methods Res [Internet]. 1981 [cited 2022 Nov 630 14];10(2):141-163. DOI https://doi.org/10.1177/00491241810100020. Available 631 at: https://journals.sagepub.com/doi/10.1177/004912418101000205 632 18. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research 633 electronic data capture (REDCap)--a metadata-driven methodology and 634 workflow process for providing translational research informatics support. J 635 Biomed Inform [Internet]. 2009 [cited 2022 Nov 14];42(2):377-381. DOI 636 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 44 10.1016/j.jbi.2008.08.010. Available at: 637 https://pubmed.ncbi.nlm.nih.gov/18929686/ 638 19. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O‘Neal L, et al. The 639 REDCap consortium: Building an international community of software platform 640 partners. J Biomed Inform [Internet]. 2019 [cited 2022 Nov 14];95:103208. DOI 641 10.1016/j.jbi.2019.103208. Available at: 642 https://pubmed.ncbi.nlm.nih.gov/31078660/ 643 20. Schönholzer TE, Pinto IC, Zacharias FCM, Gaete RAC, Serrano-Gallardo MDP. 644 Implementation of the e-SUS Primary Care system: Impact on the routine of 645 Primary Health Care professionals. Rev Lat Am Enfermagem [Internet]. 2021 646 [cited 2022 Nov 15];19(29):e3447. DOI 10.1590/1518-8345.4174.3447. 647 Available at: https://pubmed.ncbi.nlm.nih.gov/34287545/ 648 21. Brazil. Ministry of Health. e-SUS APS [Internet]. Brasília: Ministry of Health; 649 2022 [cited 2022 Nov 15]. Available at: https://sisaps.saude.gov.br/esus/ 650 22. Bardin L. Análise de Conteúdo. 2010th ed. Lisboa: Edições 70; 1977. 225 p. 651 ISBN: 978-972-44-1154-5. 652 23. Lawshe CH. A quantitative approach to content validity. Pers Psychol. 653 1975;28(4):563-575. 654 24. Baghestani AR, Ahmadi F, Tanha A, Meshkat M. Bayesian critical values for 655 Lawshe‘s content validity ratio. Meas Eval Couns Dev [Internet]. 2019 [cited 656 2022 Nov 15];52(1):69-73. DOI 10.1080/07481756.2017.1308227. Available at: 657 https://www.tandfonline.com/doi/abs/10.1080/07481756.2017.1308227 658 25. Ayre C, Scally AJ. Critical values for lawshe’s content validity ratio: revisiting 659 the original methods of calculation. Meas Eval Couns Dev [Internet]. 2014 660 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 45 [cited 2022 Nov 15];47(1):79-86. DOI 10.1177/0748175613513808. Available 661 at: https://www.tandfonline.com/doi/abs/10.1177/0748175613513808 662 26. Ortiz La Banca R, Rebustini F, Alvarenga WA, Carvalho EC, Lopes M, 663 Milaszewski K, et al. Checklists for Assessing Skills of Children With Type 1 664 Diabetes on Insulin Injection Technique. J Diabetes Sci Technol [Internet]. 2022 665 [cited 2022 Dec 21];16(3):742–750. DOI 10.1177/1932296820984771. 666 Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9294572/ 667 27. Palacio DC, Rebustini F, Oliveira DB, Peres Neto J, Barbieri W, Sanchez TP. 668 Dental vulnerability scale in primary health care: evidence of content and 669 structure internal validity. BMC Oral Health [Internet]. 2021 [cited 2022 Dec 670 21];21(1) DOI 10.1186/s12903-021-01742-6. Available from: 671 https://pubmed.ncbi.nlm.nih.gov/34454449/ 672 28. DeVellis RF. Scale development: theory and applications. 4th ed. Thousand 673 Oaks: Sage; 2017. 261 p. 674 29. PlanificaSUS. e-Planifica [Internet]. São Paulo: PlanificaSUS; 2022 [cited 2022 675 Nov 15]. Available at: https://planificasus.com.br/ 676 30. Lorenzo-Seva U, Ferrando PJ. MSA: The Forgotten Index for Identifying 677 Inappropriate Items Before Computing Exploratory Item Factor Analysis. 678 Methodology [Internet]. 2021 [cited 2022 Nov 15];17(4):296-306. DOI 679 https://doi.org/10.5964/meth.7185. Available at: 680 https://meth.psychopen.eu/index.php/meth/article/view/7185 681 31. Lorenzo-Seva U, Van Ginkel JR. Multiple Imputation of missing values in 682 exploratory factor analysis of multidimensional scales: estimating latent trait 683 scores. Anal. Psicol [Internet]. 2016 [cited 2022 Nov 15];32(2):596-608. DOI 684 https://dx.doi.org/10.6018/analesps.32.2.215161. Available at: 685 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 46 https://scielo.isciii.es/scielo.php?pid=S0212-686 97282016000200034&script=sci_abstract&tlng=en 687 32. Timmerman ME, Lorenzo-Seva U. Dimensionality assessment of ordered 688 polytomous items with parallel analysis. Psychol Methods [Internet]. 2011 [cited 689 2022 Nov 15];16(2):209-220. DOI 10.1037/a0023353. Available at: 690 https://pubmed.ncbi.nlm.nih.gov/21500916/ 691 33. Lim S, Jahng S. Determining the number of factors using parallel analysis and 692 its recent variants. Psychol Methods [Internet]. 2019 [cited 2022 Nov 693 15];24(4):452. DOI 10.1037/met0000230. Available at: 694 https://pubmed.ncbi.nlm.nih.gov/31180694/ 695 34. Auerswald M, Moshagen M. How to determine the number of factors to retain in 696 exploratory factor analysis: A comparison of extraction methods under realistic 697 conditions. Psychol Methods [Internet]. 2019 [cited 2022 Nov 15];24(4):468–698 491. DOI https://doi.org/10.1037/met0000200. Available at: 699 https://psycnet.apa.org/doiLanding?doi=10.1037%2Fmet0000200 700 35. Dobriban E, Owen AB. Deterministic parallel analysis: an improved method for 701 selecting factors and principal components. J R Stat Soc Series B Stat Methodol 702 [Internet]. 2019 [cited 2022 Nov 15];81(1):163-183. DOI 703 https://doi.org/10.1111/rssb.12301. Available at: 704 https://rss.onlinelibrary.wiley.com/doi/abs/10.1111/rssb.12301 705 36. Cho SJ, Li F, Bandalos D. Accuracy of the parallel analysis procedure with 706 polychoric correlations. Educ Psychol Meas [Internet]. 2009 [cited 2022 Nov 707 15];69(5):748-759. DOI https://doi.org/10.1177/0013164409332229. Available 708 at: https://journals.sagepub.com/doi/10.1177/0013164409332229 709 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 47 37. Briggs NE, MacCallum RC. Recovery of Weak Common Factors by Maximum 710 Likelihood and Ordinary Least Squares Estimation. Multivariate Behav Res 711 [Internet]. 2003 [cited 2022 Nov 15];38(1):25-56. DOI 712 10.1207/S15327906MBR3801_2. Available at: 713 https://pubmed.ncbi.nlm.nih.gov/26771123/ 714 38. Choi J, Kim S, Chen J, Dannels S. A Comparison of Maximum Likelihood and 715 Bayesian Estimation for Polychoric Correlation Using Monte Carlo Simulation. 716 J Educ Behav Stat [Internet]. 2011 [cited 2022 Nov 15];36(4):523–549. DOI 717 https://doi.org/10.3102/1076998610381398. Available at: 718 https://journals.sagepub.com/doi/10.3102/1076998610381398 719 39. Jo/i3re skog KG , So/i3 rbom D. LISREL V: analysis of linear structural 720 relationships by maximum likelihood and least squares methods. 5th ed. 721 Uppsala, Sweden: University of Uppsala; 1981. 722 40. Holgado-Tello FP, Chacón-Moscoso S, Barbero-García I, Vila-Abad E. 723 Polychoric versus Pearson correlations in exploratory and confirmatory factor 724 analysis of ordinal variables. Qual Quant [Internet]. 2010 [cited 2022 Nov 725 15];44(1):153–166. DOI https://doi.org/10.1007/s11135-008-9190-y. Available 726 at: https://link.springer.com/article/10.1007/s11135-008-9190-y 727 41. Baglin J. Improving your exploratory factor analysis for ordinal data: A 728 demonstration using FACTOR. Practical Assessment, Research & Evaluation 729 [Internet]. 2014 [cited 2022 Nov 15];19(5) DOI https://doi.org/10.7275/dsep-730 4220. Available at: https://scholarworks.umass.edu/pare/vol19/iss1/5/ 731 42. Osborne JW, Banjanovic ES. Exploratory factor analysis with SAS. North 732 Carolina: SAS Institute Inc.; 2016. 733 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 48 43. Lorenzo-Seva U. Promin: a method for oblique factor rotation. Multivariate 734 Behav Res [Internet]. 1999 [cited 2022 Nov 15];34(3):347-365. DOI 735 10.1207/s15327906mbr3403_3. Available at: 736 https://www.tandfonline.com/doi/abs/10.1207/S15327906MBR3403_3 737 44. Ferrando PJ, Lorenzo-Seva U. Assessing the quality and appropriateness of 738 factor solutions and factor score estimates in exploratory item factor analysis. 739 Educ Psychol Meas [Internet]. 2018 [cited 2022 Nov 15];78(5):762-780. DOI 740 10.1177/0013164417719308. Available at: 741 https://pubmed.ncbi.nlm.nih.gov/32655169/ 742 45. Hair JF, Babin BJ, Anderson RE, Black WC. Multivariate Data Analysis. 8th ed. 743 [place unknown]: CENGAGE INDIA; 2018. 832 p. 744 46. Costello AB, Osborne J. Best practices in exploratory factor analysis: Four 745 recommendations for getting the most from your analysis. Practical Assessment, 746 Research & Evaluation [Internet]. 2005 [cited 2022 Nov 15];10(1):7. DOI 747 https://doi.org/10.7275/jyj1-4868. Available at: 748 https://scholarworks.umass.edu/cgi/viewcontent.cgi?article=1156&context=pare 749 47. Wu AD, Zumbo BD, Marshall SK. A method to aid in the interpretation of EFA 750

An application of Pratt’s measures. Int J Behav Dev [Internet]. 2014 751 [cited 2022 Nov 15];38(1):98-110. DOI 752 https://doi.org/10.1177/0165025413506143. Available at: 753 https://journals.sagepub.com/doi/abs/10.1177/0165025413506143 754 48. Wu AD, Zumbo BD. Using Pratt‘s Importance Measures in Confirmatory Factor 755 Analyses. J Mod Appl Stat Methods [Internet]. 2017 [cited 2022 Nov 756 15];16(2):81-98. DOI 10.22237/jmasm/1509494700. Available at: 757 https://digitalcommons.wayne.edu/jmasm/vol16/iss2/5/ 758 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 49 49. Crobach LJ. Coefficient alpha and the internal structure of tests. Psychometrika 759 [Internet]. 1951 [cited 2022 Nov 15];16(3):297–334. DOI 760 https://doi.org/10.1007/BF02310555. Available at: 761 https://link.springer.com/article/10.1007/BF02310555 762 50. Woodhouse B, Jackson PH. Lower bounds for the reliability of the total score on 763 a test composed of non-homogeneous items: II: A search procedure to locate the 764 greatest lower bound. Psychometrika [Internet]. 1977 [cited 2022 Nov 765 15];42(4):579–591. DOI https://doi.org/10.1007/BF02295980. Available at: 766 https://link.springer.com/article/10.1007/BF02295980 767 51. McDonald RP. Test theory: A unified treatment. 1st ed. [place unknown]: 768 Psychology Press; 2013. 769 52. Ferrando PJ, Lorenzo-Seva U. A note on improving EAP trait estimation in 770 oblique factor-analytic and item response theory models. Psicologica [Internet]. 771 2016 [cited 2022 Nov 15];37:235-247. Available at: 772 https://www.uv.es/revispsi/articulos2.16/7Ferrando.pdf 773 53. Gütlein M, Helma C, Karwath A, Kramer S. A Large-Scale Empirical 774 Evaluation of Cross-Validation and External Test Set Validation in (Q)SAR. 775 Mol Inform [Internet]. 2013 [cited 2022 Nov 15];32(5-6):516-528. DOI 776 10.1002/minf.201200134. Available at: 777 https://pubmed.ncbi.nlm.nih.gov/27481669/ 778 54. Lorenzo-Seva U. SOLOMON: a method for splitting a sample into equivalent 779 subsamples in factor analysis. Behav Res Methods [Internet]. 2021 [cited 2022 780 Nov 15];:1-13. DOI https://doi.org/10.3758/s13428-021-01750-y. Available at: 781 https://link.springer.com/article/10.3758/s13428-021-01750-y 782 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 50 55. Brown TA. Confirmatory factor analysis for applied research. 2nd ed. [place 783 unknown]: Guilford publications; 2015. 462 p. ISBN: 9781462515363. 784 56. Mialhe FL, Sampaio HAC, Moraes KL, Brasil VV, Rebustini F. Psychometric 785 properties of the Brazilian version of the European Health Literacy Survey 786 Questionnaire short form. Health Promot Int [Internet]. 2022 [cited 2022 Nov 787 15];37(4):daac130. DOI 10.1093/heapro/daac130. Available at: 788 https://pubmed.ncbi.nlm.nih.gov/36102478/ 789 57. Mialhe FL, Moraes KL, Bado FMR, Brasil VV, Sampaio HAC, Rebustini F. 790 Psychometric properties of the adapted instrument European Health Literacy 791 Survey Questionnaire short-short form. Rev Lat Am Enfermagem [Internet]. 792 2021 [cited 2022 Nov 15];29:e3436. DOI 10.1590/1518-8345.4362.3436. 793 Available at: https://pubmed.ncbi.nlm.nih.gov/34231791/ 794 58. Koul A, Becchio C, Cavallo A. Cross-validation approaches for replicability in 795 psychology. Front Psychol [Internet]. 2018 [cited 2022 Nov 15];9:1117. DOI 796 10.3389/fpsyg.2018.01117. Available at: 797 https://pubmed.ncbi.nlm.nih.gov/30034353/ 798 59. Hair JR, Black WC, Babin BJ, Anderson R, Tathm RL. Multivariate Data 799 Analysis. 7th ed. Upper Saddle River: Pearson Prentice Hall; 2019. 800 60. Boedeker P, Kearns NT. Linear Discriminant Analysis for Prediction of Group 801 Membership: A User-Friendly Primer. Adv Methods Pract Psychol Sci 802 [Internet]. 2019 [cited 2022 Dec 21];2(3):250-263. DOI 803 10.1177/2515245919849378. Available from: 804 https://journals.sagepub.com/doi/full/10.1177/2515245919849378 805 61. Tabachnick BG, Fidell LS. Using multivariate statistics. Boston: Pearson; 2019. 806 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 51 62. Marmot M. Social justice, epidemiology and health inequalities. Eur J 807 Epidemiol [Internet]. 2017 [cited 2022 Nov 22];32(7):537–546. DOI 808 10.1007/s10654-017-0286-3. Available at: 809 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5570780/#CR28 810 63. Marmot M. Social determinants of health inequalities. The Lancet [Internet]. 811 2005 [cited 2022 Nov 22];365(9464):1099-1104. DOI 812 https://doi.org/10.1016/S0140-6736(05)71146-6. Available at: 813 https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(05)71146-814 6/fulltext 815 64. Dahlgren G, Whitehead M, compilers. Policies and strategies to promote social 816 equity in health [bibliography on the Internet]. [place unknown: publisher 817 unknown]; 1991 [cited 2022 Nov 22]. Available at: 818 https://core.ac.uk/download/pdf/6472456.pdf 819 65. Preda A, Voigt K. The Social Determinants of Health: Why Should We Care?. 820 Am J Bioeth [Internet]. 2015 [cited 2022 Nov 22];15(3):25-36. DOI 821 10.1080/15265161.2014.998374. Available at: 822 https://pubmed.ncbi.nlm.nih.gov/25786009/ 823 66. Barata RB, Almeida MF, Montero CV, Silva ZP. Health inequalities based on 824 ethnicity in individuals aged 15 to 64, Brazil, 1998. Cad Saude Publica 825 [Internet]. 2007 [cited 2022 Dec 21];23(2):305-313. Available from: 826 https://www.scielosp.org/article/csp/2007.v23n2/305-313/#ModalArticles 827 67. Mechanic D, Tanner J. Vulnerable people, groups, and populations: societal 828 view. Health Aff (Millwood) [Internet]. 2007 [cited 2022 Nov 22];26(5):1220-829 1230. DOI 10.1377/hlthaff.26.5.1220. Available at: 830 https://pubmed.ncbi.nlm.nih.gov/17848429/ 831 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 52 68. Mao M, Zang L, Zhang H. The Effects of Parental Absence on Children 832 Development: Evidence from Left-Behind Children in China. Int J Environ Res 833 Public Health [Internet]. 2020 [cited 2022 Nov 24];17(18):6770. DOI doi: 834 10.3390/ijerph17186770. Available at: 835 https://pubmed.ncbi.nlm.nih.gov/32957472/ 836 69. Chen L, Wulczyn F, Huhr S. Parental absence, early reading, and human capital 837 formation for rural children in China. J Community Psychol [Internet]. 2022 838 [cited 2022 Nov 24];:1-14. DOI https://doi.org/10.1002/jcop.22786. Available 839 at: https://onlinelibrary.wiley.com/doi/10.1002/jcop.22786 840 70. Fu M, Xue Y, Zhou W, Yuan T. Parental absence predicts suicide ideation 841 through emotional disorders. PLOS ONE [Internet]. 2017 [cited 2022 Nov 842 24];12(12):e0188823. DOI https://doi.org/10.1371/journal.pone.0188823. 843 Available at: 844 https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0188823 845 71. Inoue Y, Fukunaga A, Stickley A, Yzawa A, Pham TTP, Nguyen CQ, et al. 846 Association between parental absence during childhood and depressive 847 symptoms in adulthood in rural Vietnam. J Affect Disord [Internet]. 2022 [cited 848 2022 Nov 24];311:479-485. DOI https://doi.org/10.1016/j.jad.2022.05.102. 849 Available at: 850 https://www.sciencedirect.com/science/article/pii/S016503272200622X 851 72. Lacey RE, Zilanawala A, Webb E, Abell J, Bell S. Parental absence in early 852 childhood and onset of smoking and alcohol consumption before adolescence. 853 Arch Dis Child [Internet]. 2018 [cited 2022 Nov 24];103(7):691-694. DOI 854 10.1136/archdischild-2016-310444. Available at: 855 https://pubmed.ncbi.nlm.nih.gov/27831906/ 856 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 53 73. Lee BX, Donnelly PD, Cohen L, Garg S. Violence, health, and the 2030 agenda: 857 Merging evidence and implementation. J Public Health Policy [Internet]. 2016 858 [cited 2022 Nov 24];37(1):1-12. DOI 10.1057/s41271-016-0011-6. Available at: 859 https://pubmed.ncbi.nlm.nih.gov/27638239/ 860 74. Huang X, King C, McAtee J. Exposure to violence, neighborhood context, and 861 health-related outcomes in low-income urban mothers. Health & Place 862 [Internet]. 2018 [cited 2022 Nov 24];54:138-148. DOI 863 https://doi.org/10.1016/j.healthplace.2018.09.008. Available at: 864 https://www.sciencedirect.com/science/article/pii/S1353829217307165?via%3D865 ihub 866 75. Oliveira MAC, Pereira IC. Atributos essenciais da Atenção Primária e a 867 Estratégia Saúde da Família. Rev Bras Enferm [Internet]. 2013 [cited 2022 Nov 868 24];66 DOI https://doi.org/10.1590/S0034-71672013000700020. Available at: 869 https://www.scielo.br/j/reben/a/5XkBZTcLysW8fTmnXFMjC6z/?lang=pt 870 76. Streiner DL, Norman GR, Cairney J. Health Measurement Scales: A practical 871 guide to their development and use [Internet]. 5th ed. Oxford: Oxford University 872 Press; 2014 [cited 2022 Dec 21]. ISBN: 9780191765452. DOI 873 https://doi.org/10.1093/med/9780199685219.001.0001. Available from: 874 https://academic.oup.com/book/24920 875 77. Pedrosa I, Suárez-Álvarez J, García-Cueto E. Content Validity Evidences: 876 Theoretical Advances and Estimation Methods. Acción Psicológica [Internet]. 877 2014 [cited 2022 Dec 21];10(2):3-18. DOI 878 https://doi.org/10.5944/ap.10.2.11820. Available from: 879 https://revistas.uned.es/index.php/accionpsicologica/article/view/11820 880 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint 54 78. Gjersing L, Caplehorn JRM, Clausen T. Cross-cultural adaptation of research 881 instruments: language, setting, time and statistical considerations. BMC Med 882 Res Methodol [Internet]. 2010 [cited 2022 Dec 21];10:10-13. DOI 883 10.1186/1471-2288-10-13. Available from: 884 https://pubmed.ncbi.nlm.nih.gov/20144247/ 885 79. Wilson FR, Pan W, Schumsky DA. Recalculation of the Critical Values for 886 Lawshe’s Content Validity Ratio. Meas Eval Couns Dev [Internet]. 2012 [cited 887 2022 Dec 21];45(3):197-210. DOI https://doi.org/10.1177/0748175612440286. 888 Available from: 889 https://www.tandfonline.com/doi/full/10.1177/0748175612440286 890 . CC-BY 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted January 12, 2023. ; https://doi.org/10.1101/2023.01.12.23284419doi: medRxiv preprint