Self-testing for proteinuria in pregnancy: a systematic review and meta-analysis

(limit 300): 300 24

Self-testing for proteinuria may help identify preeclampsia risk during pregnancy, 25 increase end-user empowerment, and reduce burden on health systems. We conducted this 26 systematic review of the impact of proteinuria self-testing during pregnancy to expand the 27 evidence base for the World Health Organization consolidated guideline on self-care 28 interventions. 29

We comprehensively searched for articles comparing the effect of proteinuria self-30 testing with clinic-based testing among pregnant individuals receiving antenatal care on the 31 following outcomes: maternal mortality or near-miss; adverse pregnancy outcomes; eclampsia or 32 pre-eclampsia; long-term diseases; follow-up care and appropriate management; self-efficacy, 33 self-determination, autonomy, and empowerment; mental health and well-being; adverse events 34 and social harms; device-related issues; intra-uterine growth restriction; preterm birth; and 35 stillbirth or perinatal death. After abstract screening and full-text review, we systematically 36 extracted data using standardized forms and summarized the relative risks of outcomes between 37 self-testing and clinic-based testing for proteinuria. We also assessed values and preferences and 38 costs of self-testing. 39

Three publications were identified for the effectiveness review; two for 40 values/preferences, and none for the cost review, mostly from high-income countries. Overall, 41 there was no statistically significant difference between self-testing and inpatient care for 42 proteinuria among women with hypertension for any of the outcomes with data available. In 43 general, both women and providers were supportive of proteinuria self-testing because it allows a 44 greater role in self-care and fewer clinic visits, though some emphasized the need to train end-45 users for proper testing and appropriate follow-up actions. 46 Page 3

Very limited evidence suggests that self-testing for proteinuria results in 47 comparable maternal and fetal outcomes as provider-testing for hypertensive pregnant 48 individuals, and is generally acceptable to end-users and providers despite some concerns. No 49 evidence of effectiveness was available for the general pregnant population. This evidence base 50 supports its feasibility as an additional option for identifying individuals at risk of preeclampsia. 51

self-testing, proteinuria, dipstick, urinalysis, pre-eclampsia, self care 52 Systematic review registration number: PROSPERO CRD42021233845 53 Page 4

54 Preeclampsia is a significant cause of maternal and perinatal morbidity and mortality, affecting 55 2-8% of pregnancies worldwide.[1] This complication is generally diagnosed in pregnant 56 individuals who experience an onset of hypertension and subsequent proteinuria (greater than 57 300 mg in one day of protein in urine) during pregnancy.[2] About one-third of individuals who 58 experience onset of proteinuria past 20 weeks of pregnancy may ultimately contract 59 preeclampsia and have increased risk of adverse pregnancy and birth outcomes.[3, 4] Therefore, 60 monitoring urine protein levels during pregnancy serves as an important intervention in 61 achieving early diagnosis and care for preeclampsia in pregnant individuals. 62 63 Measuring proteinuria early in pregnancy can help identify individuals who are at a high risk of 64 preeclampsia and related complications, including preterm delivery and fetal malformations.[5] 65 Screening for proteinuria is typically done through dipstick urinalysis, which requires a small 66 sample of clean urine and provides a rapid result.[6] Since a recognized limitation of dipstick 67 urinalysis is poor specificity for preeclampsia versus kidney function,[7] subsequent 24-hour 68 urine collection or spot urine protein:creatinine ratio is used to verify positive proteinuria 69 findings from dipstick testing [8] as clinically appropriate. Dipstick urinalysis is typically done at 70 the point-of-care during routine antenatal care (ANC) contacts. Yet, frequency of ANC contacts 71 and proteinuria screening varies by setting, and disparities exist between settings.[9] Routine 72 screening in clinic is time-consuming and can become expensive due to frequent false positives, 73 which require further testing.[6] 74 75 Page 5 Emerging research suggests that screening can also be done through self-testing. Self-testing of 76 proteinuria may be useful for early detection and care of preeclampsia, and for reducing the 77 burden of care visits during pregnancy. Self-testing may also help pregnant individuals feel 78 involved with their care. Self-testing may also be a feasible way to promote health awareness and 79 management during health emergencies like the COVID-19 pandemic, when access to healthcare 80 services may be limited; a pulse survey conducted by the World Health Organization (WHO) on 81 the continuity of essential health services during the COVID-19 pandemic showed that ANC 82 services were among the most severely disrupted.[10] We conducted this systematic review in 83 the context of expanding the evidence base of the WHO's guideline on self-care interventions 84 [11] to include additional considerations related to maternal and perinatal health. 85 86

87 In this review, we evaluated the current literature to inform decisions on whether self-testing for 88 proteinuria during pregnancy should be available in addition to clinic-based testing. We assessed 89 three areas relevant to this topic: (1) effectiveness of the intervention, (2) values and preferences 90 of end users and providers, and (3) cost information. We followed the Preferred Reporting Items 91 for Systematic review and Meta-Analysis (PRISMA) guidelines,[12] and the protocol was 92 registered on the International Prospective Register of Systematic Reviews (PROSPERO 93 #CRD42021233845). Ethical approval was not required for this systematic review, since all data 94 came from published articles. 95 96 Effectiveness review inclusion criteria 97 Page 6 The effectiveness review was designed according to the PICO (Population, Intervention, 98 Comparison, Outcomes) format as follows: 99 100 - Population: Pregnant individuals 101 - Intervention: Self-testing for proteinuria (either by the pregnant individual or by another 102 layperson, such as a family member) 103 - Comparison: Clinic proteinuria testing by health care providers during ANC contacts 104 only 105 - Outcomes: 106 o Maternal outcomes: 107 1. Maternal mortality or near-miss 108 2. Adverse pregnancy outcomes (e.g. spontaneous abortion, premature rupture of 109 membrane, placental abruption) 110 3. Eclampsia or pre-eclampsia 111 4. Long-term (after pregnancy) cardiovascular risk, chronic hypertension, 112 diabetes, stroke 113 5. Follow-up care and appropriate management 114 6. Self-efficacy, self-determination, autonomy, and empowerment 115 7. Mental health and well-being (anxiety, stress, self-harm) 116 8. Adverse events and social harms (including discrimination, intimate partner 117 violence, stigma), and whether these harms were corrected/had redress 118 available 119 Page 7 9. Device-related issues (e.g. test failure, problems with manufacturing, 120 packaging, labeling, or instructions for use) 121 o Fetal/newborn outcomes: 122 1. Intra-uterine growth restriction 123 2. Preterm birth 124 3. Stillbirth or perinatal death 125 126 For our review, we included studies that compared self-testing for proteinuria to the comparator 127 group. The studies could be randomized controlled trials (RCTs), non-RCTs, or comparative 128 observational studies, which would include prospective controlled cohort studies, cross-sectional 129 studies, controlled before-after studies and interrupted time series, as long as they compared 130 individuals who received the intervention to those who did not. In addition, the included studies 131 were limited to peer-reviewed publications that measured one or more outcomes of interest to 132 our review. 133 134 No restrictions were placed based on the location of intervention. No language restrictions were 135 used on the search. Articles in English, French, Spanish, and Chinese were coded directly; 136 articles in other languages were translated. 137 138 Search strategy 139 We searched PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), 140 Latin American & Caribbean Health Sciences Literature (LILACS) and Embase through the 141 Page 8 search date of November 16, 2020 using the search string as follows (designed for PubMed and 142 adapted for other databases): 143 144 (proteinuria [Mesh] OR proteinuria [tiab] OR “urinary protein excretion” [tiab] OR 145 "urinalysis" [Mesh] OR "reagent strips" [Mesh] OR creatinine [Mesh] OR 146 "dipstick"[tiab]) 147 148 AND 149 150 (pregnancy [Mesh] OR pregnancy [tiab] OR pregnant [tiab] OR peri-natal [tiab] OR 151 perinatal [tiab] OR antenatal [tiab] OR maternal [tiab]) 152 153 AND 154 155 ("self care"[Mesh] OR "self-care"[tiab] OR "self-monitoring"[tiab] OR "self-156 management"[tiab] OR "self-monitor"[tiab] OR "self-manage"[tiab] OR "self-157 monitored"[tiab] OR "self-managed"[tiab] OR "self-evaluate"[tiab] OR "self-158 evaluating"[tiab] OR "self-evaluation"[tiab] OR "self-test"[tiab] OR "self-testing"[tiab] 159 OR "home"[tiab] OR "pharmacy"[tiab]) 160 161 We also searched for ongoing RCTs through clinicaltrials.gov, the WHO International Clinical 162 Trials Registry Platform, the Pan African Clinical Trials Registry, and the Australian New 163 Zealand Clinical Trials Registry. Additionally, we searched the Cochrane Library for primary 164 Page 9 research articles cited in relevant reviews. Secondary reference searching was also conducted on 165 all studies included in this review, and we asked experts in the field to help us identify additional 166 articles. 167 168 A member of the study staff screened the titles, abstracts, citation information, and descriptor 169 terms of citations that were identified through the search strategy. Full text articles were obtained 170 of all abstracts, and two independent reviewers assessed them for final study selection. 171 Differences were resolved through consensus and, if needed, with the intervention of senior staff 172 members. 173 174 Data extraction, management, and analysis 175 Data were extracted independently by two reviewers using standardized data extraction forms. 176 Differences in data extraction were resolved through consensus and referral to a senior study 177 team member from WHO when necessary. The components of information that were gathered 178 included: (1) study identification (authors, type of citation; year of publication), (2) study 179 description (study objectives, location, population characteristics, type of proteinuria monitoring 180 such as the brand of urine dipstick), (3) description of self-testing access, (4) description of any 181 additional intervention components (e.g. any education, training, support provided), (5) study 182 design, (6) sample size, (7) follow-up periods, (8) loss to follow-up, and (9) outcomes (analytic 183 approach, outcome measures, comparison groups, effect sizes, confidence intervals, significance 184 levels, conclusions, study limitations). 185 186 Page 10 Data were analyzed according to coding categories and outcomes. For each outcome assessed in 187 the review, we summarized data in Grading of Recommendations Assessment, Development and 188 Evaluation (GRADE) Evidence Profile tables using GRADEPro [15] and in a summary of 189 effects table. When multiple studies reported the same outcome, meta-analysis was conducted 190 with program Comprehensive Meta-Analysis (CMA) using random-effects models to combine 191 risk ratios. Heterogeneity was assessed using I-squared and Q statistics. We used RCT data 192 where they were available; if RCT data were not available for an outcome, we would have pulled 193 data from observational studies. For RCTs, risk of bias was assessed using the Cochrane 194 Collaboration’s tool for assessing risk of bias.[13] For non-RCTs but comparative studies, study 195 rigor would have been assessed using the Evidence Project 8-item checklist for intervention 196 evaluations.[14] 197 198 Where possible, all analyses were stratified by the following categories or subgroups: (1) 199 location or context of self-testing (ambulatory, hospitalized, or additional to standard antenatal 200 clinic contacts), (2) prior risk of preeclampsia, (3) vulnerabilities (e.g. obesity, age, poverty, 201 disability, rural/urban, literacy/educational level), and (4) high-income versus low- or middle-202 income countries. 203 204 Complementary reviews 205 The same search terms were used to search and screen for studies to be included in the values 206 and preferences and costs reviews. These studies could be qualitative or quantitative in nature, 207 but had to present primary data collection – think pieces and review articles were not included. 208 Page 11 Literature was summarized qualitatively and organized by study design and methodology, 209 location, and population. 210 211 Values and preferences review 212 Studies were included in this review if they presented primary data examining preferences of 213 individuals regarding self-testing of proteinuria during pregnancy. We focused on studies 214 examining the values and preferences of individuals who were self-testing for proteinuria during 215 pregnancy or potentially eligible for this intervention. We also included studies examining the 216 values and preferences of healthcare providers. We considered issues related to eligibility, 217 accessibility, informed decision-making, coercion, and seeking redress in this section; this 218 included the effects of stock-outs or availability of urine dipsticks. 219 220 Cost review 221 Studies were included in this review if they presented primary data comparing costing, cost-222 effectiveness, cost-utility, or cost-benefit of the intervention and comparison listed in the PICO 223 question above, or if they presented cost-effectiveness of the intervention as it related to the 224 PICO outcomes listed above. This included both cost to the health system and cost to the end-225 user. Cost literature was classified into four categories: health sector costs, other sector costs, 226 patient/family costs, and productivity impacts. 227 228 Patient and public involvement 229 Feedback on the review protocol and analysis was received from the WHO patient safety 230 working group. Patients were involved in a global survey of values and preferences conducted to 231 Page 12 inform the WHO guideline on self-care interventions and play a role in the overall 232 recommendation informed by this review. 233 234

235 Our database search yielded 398 records, and we identified another 8 through hand- and 236 secondary searching (Figure 1). Of the 334 unique records, we retained 20 for full-text review. 237 Ultimately, we included three studies in the effectiveness review, two in the values and 238 preferences review, and none in the cost review. 239 240 Figure 1. PRISMA flow chart showing disposition of citations through the search and 241 screening process 242 243 Effectiveness review 244 Three studies met the inclusion criteria for the effectiveness review. These were small RCTs 245 (sample sizes ranged from n=63 to n=218) from the United Kingdom (UK), Zimbabwe, and 246 Hong Kong taking place over 20 years ago, comparing the effectiveness of antenatal ward 247 admission (including proteinuria testing at the hospital) versus home-based standard of care 248 (including proteinuria self-testing using dipsticks) among pregnant participants with varying 249 degrees of non-proteinuric hypertension.[15-17] In the control (self-testing) arm of the included 250 studies, which is the intervention for this review, pregnant women were given instructions on 251 how to self-test for proteinuria, were provided an instruction sheet and container of urine dip 252 sticks to take home, and were encouraged to test every day. They were advised to continue 253 normal activity at home with no particular restrictions and received regular ANC at the local day 254 Page 13 care center or outpatient clinic either weekly [15, 16] or once every two weeks [17]. These 255 participants were given instructions to contact the health facility for a consult or to be admitted if 256 the dipsticks indicated proteinuria, among other health conditions (e.g. other signs/symptoms of 257 pre-eclampsia like increasing headache or abdominal pain, indicators of labor, no fetal 258 movement). In the intervention arm of the included trials, pregnant participants were admitted to 259 the antenatal ward after study recruitment and received daily examination by health providers 260 (e.g. obstetricians, midwives), blood pressure monitoring, and urine proteinuria testing. Activity 261 (while an inpatient or on bed rest) was limited, though voluntary ambulation (e.g. to meals and 262 toilet) was allowed. 263 264 For the purposes of our review, the relative risks of outcomes were recalculated to treat self-265 testing at home as the intervention and clinic/hospital-based testing as the control. Study 266 descriptions are presented in Table 1 and the summary of effects in Table 2. It was not possible 267 to further stratify the outcome data given the small number of studies. All RCTs reported on the 268 development of maternal severe hypertension, birthweight in grams, small-for-gestational age, 269 and preterm birth. Low birthweight was reported in the Zimbabwe study and Hong Kong 270 study.[15, 16] The UK study reported on eclampsia, albuminuria, stillbirth, and neonatal 271 mortality;[17] the Hong Kong study reported on maternal development of proteinuria and 272 development of severe proteinuria.[16] In general, the certainty of evidence for the reported 273 outcomes was low to very low because of indirectness (the comparator – hospital admission – 274 went beyond our desired clinic-based proteinuria testing during ANC, and the intervention – 275 home-based normal activity with daily proteinuria testing – included more components than 276 Page 14 solely proteinuria self-testing) and imprecision (very small sample sizes and very low event 277 rates). 278 Page 15 Table 1. Description of included studies. Study Location Population Sampling Intervention Comparator Mathews 1977 RCT United Kingdom: Kent Rural Pregnant women (>28 weeks gestation, singleton) with diastolic blood pressure 90- 109 mm Hg after 5 minutes' rest, without sedation N=28 (self); 35 (provider) Non- probability facility-based Daily self-testing for urine proteinuria testing at home via Albustix, as part of "routine outpatient care" (reviewed biweekly until 36 weeks, every week thereafter) and "normal activity at home" Proteinuria testing by the nurse/physician at the clinic, as part of “inpatient bed rest in hospital without sedation” Crowther 1992 RCT Zimbabwe: Harare Urban Pregnant women (28-38 weeks gestation, singleton) with diastolic blood pressure 90- 110 mm Hg but no/trace proteinuria N=108 (self); 110 (provider) Non- probability facility-based Daily self-testing for urine proteinuria at home via Albustix, as part of "routine outpatient care" (reviewed weekly) and "normal activity at home" Proteinuria testing on a daily basis by the nurse/physician at the clinic, as part of "admission to hospital for bed rest" Leung 1998 RCT Hong Kong Urban Pregnant women (28-38 weeks gestation, singleton) with diastolic blood pressure 90- 100 mm Hg after 5 minutes' rest N=44 (self); 44 (provider) Non- probability facility-based Daily self-testing urine proteinuria at home via Albustix, as part of routine care at a day care clinic or outpatient care and "normal activity at home" Proteinuria testing on a daily basis by the nurse/physician at the clinic, as part of "inpatient admission to antenatal ward for bed rest" Page 16 Table 2. Summary of effects from meta-analyses (number of studies n>1) or from single studies. Outcome Study type Number of studies RR 95% CI p-value for RR Q- value p-value for Q statistic I- squared Maternal Outcomes Eclampsia or pre-eclampsia Eclampsia RCT 1 No events reported in either arm Long-term cardiovascular risk, chronic hypertension, diabetes, stroke Albuminuria RCT 1 0.63 0.12 to 3.17 0.57 NA NA NA Development of severe hypertension (assessed with: diastolic blood pressure >109 mm Hg) RCT 3 0.91 0.30 to 2.70 0.86 5.28 0.07 62.10 Development of proteinuric hypertension and DBP 1+ on albustix testing) RCT 1 1.02 0.83 to 1.25 0.86 NA NA NA Development of severe proteinuria (assessed with: >3+ on albustix testing) RCT 1 1.44 0.92 to 2.26 0.11 NA NA NA Fetal/Newborn Outcomes Intrauterine growth restriction Birthweight in grams (difference in means) RCT 3 -20.96 -134.51 to 92.58 0.72 1.29 0.53 0 Small for gestational age (assessed with: birthweight <10th percentile) RCT 3 0.91 0.52 to 1.58 0.74 1.97 0.37 0 Low birthweight (assessed with: birthweight <2500 grams) RCT 2 1.52 0.79 to 2.93 0.21 0.03 0.87 0 Preterm birth Preterm birth (assessed with: delivery at gestational age <37 weeks) RCT 3 1.66 0.95 to 2.91 0.08 1.15 0.56 0 Stillbirth or perinatal death Stillbirth (assessed with: fetal death before onset of labor) RCT 1 No events reported in either arm Neonatal mortality (assessed with: number of deaths at 0-28 days of birth) RCT 1 No events reported in either arm Page 17 Eclampsia or pre-eclampsia Though one RCT measured eclampsia as an outcome, no events of eclampsia were reported in either study arm.[17] This was graded as very low certainty evidence that self-testing for proteinuria had no difference compared to clinic-based testing on eclampsia or pre-eclampsia. Long-term cardiovascular risk, chronic hypertension, diabetes, stroke Three RCTs included indirect measures of long-term cardiovascular risk, chronic hypertension, diabetes, and stroke.[15-17] Meta-analysis of three RCTs found no difference in the risk of developing severe hypertension between pregnant women who self-tested for proteinuria and those who tested in clinics (RR: 0.91, 95% CI: 0.30-2.70, I-squared: 62.10). Between pregnant women who self-tested for albuminuria and those who were tested at a clinic, one RCT found no statistically significant difference in the risk of developing albuminuria (RR: 0.63, 95% CI: 0.12- 3.17).[17] One RCT found no difference in the risk of developing proteinuric hypertension when comparing pregnant women who self-tested for proteinuria to pregnant women at the clinic (RR: 1.29, 95% CI: 0.23-7.24).[16] The same RCT found no difference between the self-testing and clinic-based testing arms in the risk of the pregnant women developing proteinuria (RR: 1.02, 95% CI: 0.83-1.25) or severe proteinuria (RR: 1.44, 95% CI: 0.92-2.26).[16] These outcomes were graded as low to very low certainty evidence showing that self-testing for proteinuria is comparable to clinic-based testing on long-term maternal health outcomes. Intra-uterine growth restriction The outcome of intra-uterine growth restriction was measured indirectly through three reported neonatal outcomes: birthweight in grams, low birthweight, and small for gestational age. Meta- Page 18 analysis of three RCTs found no difference in birthweight between self-testing and clinic-based testing for proteinuria (MD: -20.96g, 95% CI: -134.51-92.58, I-squared: 0).[15-17] Meta- analysis of two RCTs showed no impact of proteinuria self-testing on the risk of low birthweight (1.52, 95% CI: 0.79-2.93, I-squared: 0).[15, 16] Meta-analysis of three RCTs found no difference in the risk of infants being born small-for-gestational-age (RR: 0.91, 95% CI: 0.52-1.58, I- squared: 0).[15-17] This was graded as low to very low certainty evidence that self-testing for proteinuria had no difference from clinic-based testing on intra-uterine growth restriction. Preterm birth Meta-analysis of three RCTs found an elevated but not statistically significant rate of preterm birth among women who self-tested for proteinuria compared to those who tested at the clinic (RR: 1.66, 95% CI: 0.95-2.91, I-squared: 0).[15-17] This was graded as very low certainty evidence that self-testing for proteinuria had no difference compared to clinic-based testing on preterm birth. Stillbirth or perinatal death One RCT measured stillbirth and neonatal mortality as outcomes; however, no events for either outcome were reported in either study arm.[17] This was graded as low to very low certainty evidence that self-testing for proteinuria had no difference compared to clinic-based testing on stillbirth or perinatal death. Other outcomes of interest Page 19 Within the included studies, no comparative data were found for the following outcomes: maternal mortality or near-miss; adverse pregnancy outcomes; follow-up care and appropriate management; self-efficacy, self-determination, autonomy, and empowerment; mental health and well-being; adverse events and social harms, and whether these harms were corrected/had redress available; or device-related issues. Values and Preferences Review Two quantitative feasibility studies for self-testing urine for proteinuria during pregnancy, one from the UK [18] and the other from the United States [19], found that most pregnant individuals were highly satisfied with or preferred self-testing for proteinuria over in-clinic testing. The common reason for liking self-testing for proteinuria across the two studies was ease of use. One study explored values and preferences of end-users and providers in greater depth.[18] In this study, at least 95% of the surveyed women expressed willingness to self-test or discuss the

with their providers. All agreed that self-testing would provide them with a sense of greater involvement in their pregnancy care, and many appreciated the reassurance they felt when they received negative test results. Although some concerns were raised about increased stress or anxiety and accurate use of the dipsticks, almost all of the women who had not self- tested previously were open to the idea if provided the necessary training, reassurance, or second opinions from providers. These concerns were generally lower amongst the survey participants who had previously self-tested for proteinuria. In addition, these individuals felt that self-testing was empowering and they liked not making unnecessary trips to the hospital. Page 20 The majority of surveyed providers saw self-testing for proteinuria as a way for women to detect pre-eclampsia early, empower themselves, and save time and money.[18] Close to 80% believed that self-testing would enhance usual care, though about 70% also reported that they would repeat urinalysis despite having women self-test. However, providers also raised concerns about pregnant individuals’ aptitude and suitability for self-testing, their ability to act appropriately on any positive results, and whether self-testing might increase demand for urgent clinic-based services. Cost Review No studies were identified for the cost review.

Among pregnant individuals with varying levels of non-proteinuric hypertension, three RCTs found no difference between self-testing and clinic-based testing for proteinuria on the risk of any of the maternal or neonatal outcomes for which data were available. These trials focused on individuals with an existing diagnosis of hypertension and compared home versus hospital management; we were unable to find any trials that compared self-testing to clinic-based testing for proteinuria during ANC contacts. We also found no comparative data for our other maternal outcomes of interest. In terms of values and preferences, most individuals found self-testing for proteinuria acceptable. Though some had concerns over their ability to accurately perform and interpret the tests, these fears were generally attenuated if training was provided. The sense of self-empowerment, ownership of care, and decreased frequency of clinic visits were also appreciated by end-users. Self-testing was positively regarded by providers for similar reasons. Page 21 Providers generally approved adding proteinuria self-testing as a supplemental option to usual ANC, despite some concerns about end-users’ ability to perform the tests correctly or interpret/follow-up appropriately on the test results. There was no data available regarding cost savings related to self-testing or clinic-based testing, either for individuals or health system. However, even where urine testing strips are available over-the-counter through pharmacies or recommended by health care providers, their cost may not be covered by health insurance and may need to be out-of-pocket for many pregnant individuals. Cost is therefore an important consideration for an intervention which may need to be conducted over several months during pregnancy. Several studies have examined the validity of self-testing for proteinuria, compared with clinic- based testing. A prospective observational study from Australia found that pregnant individuals who self-tested for proteinuria with dipstick tended to overestimate protein levels in their urine.[20] Another prospective cohort study found that the dipstick test performance was similar to the other proteinuria assessment methods that are typically used in the clinic setting, including protein:creatine ratio and 24-hour collection checks.[21] Even so, controversy surrounding the clinical utility of urine dipstick testing (including self- testing) persists, in part due to the lack of a gold standard for diagnosis of proteinuria in pregnancy. Urine dipstick as an indicator for proteinuria is subject to several limitations, including variability in urine concentration depending on fluid status, the time of the day during which the test takes place, and whether they had urinated prior to testing. A diagnosis of proteinuria may require, at minimum, the indicator value of +2 on a dipstick, which should still Page 22 be confirmed by a quantitative test.[22] Considering this lack of specificity, many providers prefer using spot urine protein-to-creatinine ratio for proteinuria testing because of its relatively high accuracy, reliability, and reproducibility.[23] Spot urine testing may be inferior to a 24-hour urine collection test due to the same limitations of dipstick urinalysis: high within-participant variability in urine protein excretion even among providers.[24] However, 24-hour urine collection is limited by the risk of contamination, over-collection, and inconvenience. Despite recognized limitations, initial proteinuria testing (whether in clinic or at home) based on dipstick urinalysis with follow-up tests as indicated may help appropriately triage patients in resource-limited settings,[25] though the value-added of routine proteinuria testing via urine dipsticks may be limited in well-resourced settings. This testing method remains a standard tool for proteinuria testing in low-resource settings where testing affordability is a key issue. With the wide variety of commercially available dipsticks, there are some affordable and effective options.[26, 27] In low-resource settings, clinic-administered urinalysis which requires access to laboratory resources is more limited than other methods to screen for preeclampsia such as blood pressure monitoring,[27] suggesting that self-testing for proteinuria may still have value to help guide appropriate level of care. Further studies will be needed to adequately assess this question. Clinical guidelines [28] highlight the need for information beyond proteinuria to diagnose and manage hypertensive complications of pregnancy, as non-proteinuric hypertensive disease is a recognized entity that has outcomes similar to preeclampsia. Though measuring proteinuria early in pregnancy can help predict individuals at high risk for important complications like preeclampsia and preterm delivery,[5] additional tests are needed for accurate assessment of Page 23 maternal health status. Other indicators of preeclampsia including blood pressure, glomerular filtration rate, and neurological signs may be more important than proteinuria in predicting adverse pregnancy outcomes.[2, 29, 30] This review has several strengths. We conducted a comprehensive screen across multiple databases as well as a hand search and secondary search, leaving little room for missing any relevant articles. We also examined the methodological quality of studies and assessed not only the effectiveness of self-testing for proteinuria but also the values and preferences of the pregnant end-users and providers. We found that proteinuria self-testing among pregnant individuals with diagnosed hypertension may be just as effective as proteinuria testing during inpatient care in predicting a wide range of maternal and neonatal outcomes, and that there is high acceptance of self-testing among both the pregnant individuals and providers who were surveyed. It is important to interpret our findings within the context of limited literature. The studies identified for the effectiveness review were conducted among pregnant individuals diagnosed with some form of hypertension, which is one of the risk factors for preeclampsia and eclampsia, not the broader pregnant population. These effectiveness studies compared self-testing to inpatient admission, which is more intensive and less comparable than clinic-based testing. While we meta-analyzed the outcome of severe hypertension, we integrated only three small RCTs, so the meta-analysis may provide a false sense of certainty that the pooled estimate represents the true effect. Because these three studies were conducted decades ago, it may be challenging to apply their findings in current clinical practice. The scope of this review was Page 24 limited to self-testing for proteinuria rather than any other of the many methods or biomarkers for identifying pregnancy complications, resulting in a small pool of studies meeting our inclusion criteria. We also found no peer-reviewed evidence on the costs related to proteinuria self-testing, although dipstick testing is likely substantially cheaper than hospitalization. Nonetheless, the currently available albeit limited data suggests that self-testing for proteinuria is not harmful. Further studies will be needed to assess whether self-testing for proteinuria as part of routine antenatal care can improve pregnancy outcomes among the general population, considering factors such as previous history of preeclampsia or eclampsia, age, obesity, race/ethnicity, and multiple pregnancies.

Very limited evidence suggests that self-testing for proteinuria yields comparable maternal and fetal outcomes as provider testing, and is generally acceptable to end-users and providers. This evidence supports its feasibility as an additional option for identifying individuals at risk of preeclampsia. List of abbreviations ANC: antenatal care CINAHL: Cumulative Index to Nursing and Allied Health Literature CMA: Comprehensive Meta-Analysis GRADE: Grading of Recommendations Assessment, Development and Evaluation LILACS: Latin American & Caribbean Health Sciences Literature PICO: Population, Intervention, Comparison, Outcomes Page 25 PRISMA: Preferred Reporting Items for Systematic review and Meta-Analysis PROSPERO: Prospective Register of Systematic Reviews RCT: randomized controlled trial UK: United Kingdom WHO: World Health Organization Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Availability of data and materials All data generated or analysed during this study are included in this published article and come from other published articles cited as included studies in this review. Competing interests The authors declare that they have no competing interests. Funding We gratefully acknowledge financial support of The Children's Investment Fund Foundation (CIFF). The funder played no part in the decision to submit the article for publication, nor in the Page 26 collection, analysis and interpretation of data. All authors had full access to all of the data in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. Authors' contributions MN conceptualized the study following input from OT. CEK and PTY designed the protocol with feedback from OT, BL, and MN. PTY ran the database search and oversaw search, screening, full text review, and data abstraction processes with support from DR. CEK and PTY performed data analysis. PTY and DR drafted the manuscript. PTY, DR, CEK, CAZ, OT, BL, and MN reviewed the draft, provided critical review, and read and approved the final manuscript. The corresponding author, as guarantor, accepts full responsibility for the finished article has access to any data and controlled the decision to publish. The corresponding author attests that all listed authors meet the authorship criteria and that no others meeting the criteria have been omitted. The named authors alone are responsible for the views expressed in this publication and do not necessarily represent the decisions or the policies of the World Health Organization (WHO) nor the UNDP-UNFPA-UNICEF-WHO-World Bank Special Programme of Research, Development and Research Training in Human Reproduction (HRP).

We thank Maurice Bucagu, Laura Ferguson, Rudolfo Gomez, Oleg Kuzmenko, and Karima Gholbzouri for their feedback on the review protocol. We also thank our Johns Hopkins graduate research assistants (Huneid Kautsar, Jaime Marquis, and Cynthia Li) for their crucial help in searching, screening, and extracting data. Page 27

1. Khan KS, Wojdyla D, Say L, Gülmezoglu AM, Van Look PFA. WHO analysis of causes of maternal death: a systematic review. The Lancet. 2006;367(9516):1066-74. 2. Fishel Bartal M, Lindheimer MD, Sibai BM. Proteinuria during pregnancy: definition, pathophysiology, methodology, and clinical significance. American Journal of Obstetrics & Gynecology. 2020. 3. Sarno L, Maruotti GM, Saccone G, Sirico A, Mazzarelli LL, Martinelli P. Pregnancy outcome in proteinuria-onset and hypertension-onset preeclampsia. Hypertension in pregnancy. 2015;34(3):284-90. 4. Ekiz A, Kaya B, Polat I, Avci ME, Ozkose B, Kicik Caliskan R, et al. The outcome of pregnancy with new onset proteinuria without hypertension: retrospective observational study. The journal of maternal-fetal & neonatal medicine : the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2016;29(11):1765-9. 5. Zhang JJ, Ma XX, Hao L, Liu LJ, Lv JC, Zhang H. A Systematic Review and Meta- Analysis of Outcomes of Pregnancy in CKD and CKD Outcomes in Pregnancy. Clinical journal of the American Society of Nephrology : CJASN. 2015;10(11):1964-78. 6. Waugh JJ, Clark TJ, Divakaran TG, Khan KS, Kilby MD. Accuracy of urinalysis dipstick techniques in predicting significant proteinuria in pregnancy. Obstetrics and gynecology. 2004;103(4):769-77. Page 28 7. Airoldi J, Weinstein L. Clinical significance of proteinuria in pregnancy. Obstetrical & gynecological survey. 2007;62(2):117-24. 8. Milne F, Redman C, Walker J, Baker P, Black R, Blincowe J, et al. Assessing the onset of pre-eclampsia in the hospital day unit: summary of the pre-eclampsia guideline (PRECOG II). Bmj. 2009;339. 9. Salam RA, Das JK, Ali A, Bhaumik S, Lassi ZS. Diagnosis and management of preeclampsia in community settings in low and middle-income countries. J Family Med Prim Care. 2015;4(4):501-6. 10. WHO. Pulse survey on continuity of essential health services during the COVID-19 pandemic: interim report, 27 August 2020. Geneva, Switzerland: World Health Organization; 2020. 11. World Health Organization. WHO consolidated guideline on self-care interventions for health: sexual and reproductive health and rights. Geneva, Switzerland: WHO; 2019. 12. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS medicine. 2009;6(7):e1000097. 13. Cochrane Handbook for Systematic Reviews of Interventions 2019. Available from: www.training.cochrane.org/handbook . 14. Kennedy CE, Fonner VA, Armstrong KA, Denison JA, Yeh PT, O’Reilly KR, et al. The Evidence Project risk of bias tool: assessing study rigor for both randomized and non- randomized intervention studies. Systematic Reviews. 2019;8(1):3. 15. Crowther CA, Bouwmeester AM, Ashurst HM. Does admission to hospital for bed rest prevent disease progression or improve fetal outcome in pregnancy complicated by non- Page 29 proteinuric hypertension? BJOG: An International Journal of Obstetrics & Gynaecology. 1992;99(1):13-7. 16. Leung KY, Sum TK, Tse CY, Law KW, Chan MY. Is in-patient management of diastolic blood pressure between 90 and 100 mm Hg during pregnancy necessary? Hong Kong medical journal = Xianggang yi xue za zhi. 1998;4(2):211-7. 17. Mathews DD. A randomized controlled trial of bed rest and sedation or normal activity and non-sedation in the management of non-albuminuric hypertension in late pregnancy. British journal of obstetrics and gynaecology. 1977;84(2):108-14. 18. Tucker KL, Bowen L, Crawford C, Mallon P, Hinton L, Lee MM, et al. The feasibility and acceptability of self-testing for proteinuria during pregnancy: A mixed methods approach. Pregnancy hypertension. 2018;12:161-8. 19. Burke AE, Thaler KM, Geva M, Adiri Y. Feasibility and acceptability of home use of a smartphone-based urine testing application among women in prenatal care. American journal of obstetrics and gynecology. 2019;221(5):527-8. 20. Goh JT, Krause H. A prospective observational study on the accuracy of patient self- testing of urine at an antenatal clinic. The Australian & New Zealand journal of obstetrics & gynaecology. 2002;42(1):67-8. 21. Payne B, Magee LA, Côté AM, Hutcheon JA, Li J, Kyle PM, et al. PIERS proteinuria: relationship with adverse maternal and perinatal outcome. Journal of obstetrics and gynaecology Canada : JOGC = Journal d'obstetrique et gynecologie du Canada : JOGC. 2011;33(6):588-97. 22. Gangaram R, Ojwang PJ, Moodley J, Maharaj D. The accuracy of urine dipsticks as a screening test for proteinuria in hypertensive disorders of pregnancy. Hypertension in pregnancy. 2005;24(2):117-23. Page 30 23. Morris RK, Riley RD, Doug M, Deeks JJ, Kilby MD. Diagnostic accuracy of spot urinary protein and albumin to creatinine ratios for detection of significant proteinuria or adverse pregnancy outcome in patients with suspected pre-eclampsia: systematic review and meta- analysis. Bmj. 2012;345:e4342. 24. Ayoub I, Williams J, Birmingham D, Hebert L. Spot Urine Protein/Creatinine Ratio Testing at a Large University Medical Center: Evidence for Overuse of This Low-Value Diagnostic Test. Kidney Int Rep. 2020;5(5):566-8. 25. Brown MA, Magee LA, Kenny LC, Karumanchi SA, McCarthy FP, Saito S, et al. Hypertensive Disorders of Pregnancy: ISSHP Classification, Diagnosis, and Management Recommendations for International Practice. Hypertension (Dallas, Tex : 1979). 2018;72(1):24- 43. 26. McLaren ZM, Sharp A, Hessburg JP, Sarvestani AS, Parker E, Akazili J, et al. Cost effectiveness of medical devices to diagnose pre-eclampsia in low-resource settings. Development Engineering. 2017;2:99-106. 27. Firoz T, Sanghvi H, Merialdi M, von Dadelszen P. Pre-eclampsia in low and middle income countries. Best practice & research Clinical obstetrics & gynaecology. 2011;25(4):537- 48. 28. Gestational Hypertension and Preeclampsia: ACOG Practice Bulletin Summary, Number 222. Obstetrics and gynecology. 2020;135(6):1492-5. 29. Lindheimer MD, Kanter D. Interpreting abnormal proteinuria in pregnancy: the need for a more pathophysiological approach. Obstetrics and gynecology. 2010;115(2 Pt 1):365-75. Page 31 30. Homer CS, Brown MA, Mangos G, Davis GK. Non-proteinuric pre-eclampsia: a novel risk indicator in women with gestational hypertension. Journal of hypertension. 2008;26(2):295- 302.