Pregnancy Outcomes and Documentation Status Among Latina Women: A Systematic Review

Purpose: The impression that Latinas experience paradoxically good pregnancy outcomes in the United States persists, despite evidence showing that these outcomes are not enjoyed by all Latina subgroups. We conducted this systematic literature review to examine the relationship between documentation status and pregnancy outcomes among Latinas. Methods: This review synthesizes empirical evidence on this relationship; examines how these studies deﬁne and operationalize documentation status; and makes recommendations of how a more comprehensive methodological approach can guide public health research on the impact of documentation status on Latina immigrants to the United States. We searched the literature within PubMed, Web of Science, Academic Search Premier, and Google Scholar in 2017 for relevant studies. Results: Based on stringent inclusion criteria, we retained nine studies for analysis. Conclusion: We found that evidence for the impact of documentation status on pregnancy outcomes among Latinas is not conclusive. We believe the divergence in our ﬁndings is, in part, due to variation in: conceptualization of how documentation status impacts pregnancy outcomes, sample populations, deﬁnitions of exposures and outcomes, and contextual factors included in models. Speciﬁc analytic challenges around sampling, measurement, and data analysis are identiﬁed. Suggestions for future research are offered regarding measurement of documentation status. Findings highlight the need for increased attention to documentation as an inﬂuence on Latina pregnancy outcomes.


Introduction
Compared with other racial and ethnic groups in the United States, Latinas* have less education, lower socioeconomic status, less access to medical care, and lower use of prenatal care [1][2][3] ; despite these risk factors, Latinas in the United States have surprisingly favorable pregnancy outcomes. This well-known phenomenon is the Latina paradox, [4][5][6] and there is substantial evidence to support its existence and impact. 7,8 Of critical note, with more time spent living in the United States, these paradoxically good pregnancy outcomes decline, and Latina health status draws closer to and sometimes below that of non-Latina Whites. [9][10][11][12] Further highlighting the inequities associated with this phenomenon, the paradox has not been demonstrated or sufficiently explored across all pregnancy outcomes or stratified by documentation status. Due to the paradox, an erroneous perception has persisted that among women of color, Latina birth outcomes are not a pressing concern. Because of this prevailing view that all U.S.based Latinas are experiencing above-optimal pregnancy outcomes (when this may not be the case), it is important to examine the paradox for variation across diverse outcomes and subgroups. Clarifying where, for whom, when, and how the paradox applies has critical implications for health equity.
Most research on the paradox has focused on low birthweight (LBW) and infant mortality (IM), finding that compared with infants of non-Latina White women, Latina infants are less likely to experience LBW 13,14 and IM. 8 But these are not the only outcomes of importance for Latinas and their offspring. Preclampsia, which places women at increased risk of maternal and fetal death 15 and has implications for adverse vascular health across the life course, 16 is more likely among Latinas than non-Latina white women 17 ; similarly, Latinas-again compared with non-Latina white women-are at greater risk of hypertension, 17 which means, among other health risks, increased risk of chronic kidney disease later in life. 18 Further, Latinas are more likely to develop gestational diabetes mellitus (GDM), a pregnancy outcome associated with prepregnancy obesity 19 and a risk factor for developing type II diabetes. 20 In fact, half of all Latina women begin pregnancy while being either overweight or obese and experience inappropriate weight gain-both inadequate and excessive 19 -making gestational weight gain (GWG) another pregnancy outcome with nonparadoxical patterns and health implications across the life course. With the exception of women who entered pregnancy underweight, 21 Latina women are more likely to report excessive GWG when compared with both Black and non-Black non-Latina women. 22 This high burden of GDM and excessive GWG among Latina women places them at increased risk of giving birth to large for gestational age (LGA) infants. 23 However, despite the Latina paradox focus on birthweight, measures of birthweight that incorporate gestational age-such as LGA and small for gestational age (SGA)-are not typically considered.
It is also notable that the paradox is not borne out across all Latina subgroups. The paradox appears to have a differential impact by nativity, with Mexicanborn women experiencing better outcomes than, for example, Central or South American women. 4,8,[24][25][26][27] It is also most strongly observed among foreign-born Latinas, despite their risk profile, including higher rates of poverty and lower levels of education. 14,28 Given the importance of nativity and nationality, a consideration of documentation status is warranted based on its impacts on immigrant well-being [29][30][31][32] ; upward mobility 29,30,32 ; and access to health care coverage [33][34][35] and utilization. 36 Immigration itself is a social determinant of health, and the social, political, and economic drivers of immigration and contexts of reception result in stratification with critical impacts on immigrant health across the lifecourse. 37 Latino immigrants have encountered an increasingly hostile context of reception 38 marked by structurally racist documentation barriers 38 and antiimmigration policies, potentially amplifying the impact of documentation status on Latina pregnancy outcomes. Community-level factors, including social networks and social support, 39,40 have also been pointed to as critical for Latina pregnancy health; this emphasis on social connection posits that these relationships among first-generation Latinas and the loss of these ties among second-generation Latinas (and beyond) explain the diminished pregnancy outcomes across time in the United States. These findings add to an emerging literature attempting to differentiate first-from secondgeneration Latina experiences. One study in this area showed that Latina immigrants experience isolation and ''othering'' as a result of structural and personally mediated racism 41 ; another demonstrated the adverse impacts of neighborhood-level poverty and ethnic density on social processes among second-generation Latinas. 42 At the individual level, acculturation and assimilation 43 processes have pointed to how immigration behaviors perceived to be culturally related may shift with years of residence in the United States (possibly associated with documentation status). Immigration stress 44 has also been pointed to as a determinant of Latina pregnancy outcomes, with stress closely linked with adverse birth outcomes. And finally, for individual Latinas, documentation status could result in differential access to health-promoting resources, since being undocumented is a known barrier for Latina immigrants in accessing prenatal care. [45][46][47] In is notable that documentation status remains relatively unexplored in the research on maternal child health inequities. There are a number of reasons why research on the impacts of documentation status is limited. Concerns about a ''chilling effect'' among participants, manifested as reluctance to participate or fear-based dishonesty about status, have resulted in persistent hesitance by researchers to collect this information in survey-based research. [48][49][50][51] The recognition that these data are sensitive and that gathering them have implications for harm 49,52 has also contributed in the following manner: Collecting, storing, and analyzing documentation status (and disseminating the results) in the context of current U.S. immigrant enforcement policies places potential research participants at great risk of discovery, detention, and deportation. 53 This barrier to scientific inquiry on the role of documentation status underscores the critical need for research aimed at understanding the relationships between being undocumented and maternal/child health. Given (1) the dearth of published research on the impact of documentation status on pregnancy outcomes; (2) our current knowledge about the inequities in outcomes across Latina subgroups; and (3) the increasingly hostile context of reception encountered by Latina immigrants to the United States, it is vital to understand the differential impacts across Latinas, specifically in documented versus undocumented women, to best meet the needs of diverse subgroups.
This systematic literature review aims to contribute to the literature by attempting to enhance our understanding of the Latina paradox by critically examining the current empirical evidence to explore how documentation status is measured and may be theorized to impact pregnancy outcomes among this population. We hypothesize that documentation status will impact pregnancy outcomes such that legal status (among foreign-born Latinas) will be protective for pregnancy outcomes (and being undocumented will increase risk for adverse outcomes). We specify this among foreign-born Latinas, because we know that U.S.-born Latinas (despite having legal status) are more likely to have worse pregnancy outcomes. This examination will further elucidate how Latinas' vulnerability to adverse outcomes is shaped and reified by documentation status. To achieve our aim, this review has three objectives: to (1) synthesize the empirical evidence on the relationship between documentation status and pregnancy outcomes among Latina women in the United States; (2) examine how these studies define and operationalize documentation status in this context; and (3) make recommendations of how a more comprehensive methodological approach can guide public health research on the impact of documentation status on Latina immigrants to the United States
We included English language published studies, white papers, reports, dissertations, and other literature detailing original observational research conducted in the United States. Studies were included if they: (1) included and/or restricted their study sample to Latina women; (2) quantitatively examined associations between documentation status and pregnancy outcomes; and (3) focused on Latina women from non-U.S. territories (due to our specific interest in the measurement and impact of documentation status).

Study selection and data extraction
As shown in Figure 1, the search process yielded an initial set of 1924 unique articles. Of this initial article set, 1444 were excluded based on title and abstract review, leaving 480 articles for full text review. Of those, six articles met our inclusion criteria. A review of these articles' reference lists yielded three additional articles, bringing the total for inclusion to nine.
Each paper identified in our search was independently examined by two authors. Paper titles were reviewed and excluded if they were clearly outside the review topic. If the title did not provide sufficient information to determine inclusion status, the abstract and subsequently the full text were reviewed. In the case of discrepant reviews, a third author examined the paper to determine inclusion/exclusion. Finally, this same process was applied to our review of the reference lists of the included papers.
Each author independently extracted information pertaining to the study design and analysis. To guide our review, we used the PRISMA reporting checklist, adapted as a Qualtrics abstraction form to facilitate capturing characteristics from each article, including: documentation status measurement; pregnancy outcomes definition and ascertainment; race/ethnicity and country of origin of study sample; covariates; and statistical approach, including management of missing data. To assess each included study's resiliency from bias, we used a modified version of the NIH Quality Assessment Tool for Observational Cohort and Crosssectional Studies (Appendix A1), with two authors independently appraising each study. Given that one purpose of this review is to report the quality of research in this area and make recommendations for future research, we include all studies in this review-irrespective of resiliency from bias-as is consistent with the emerging nature of this research topic.
This study was exempted by the Portland State University institutional review board.

Results
Of the 1924 articles initially identified through our review process, only 6 met our full inclusion criteria; another 3 articles found through reference checks brought the total reviewed articles to 9. Exclusion of abstracts was primarily due to study populations not inclusive of Latinas and/or not capturing pregnancy outcomes. Exclusion of full-text articles resulted when the study did not specify and measure documentation status and/or pregnancy outcomes were limited to adequacy of prenatal care.
Studies examined nine unique cohorts of women and their infants born from 1980 through 2008, utilizing birth records, claims data, and/or in-person interviews to procure data. For all but two studies, 54,55 outcomes were ascertained via administrative data or medical record extraction. These two studies captured outcome measures directly from participants: The first 54 ascertained information about cesarean deliveries via self-report, and the second 55 assessed postpartum depression with the Center for Epidemiologic Studies Depression scale. The majority of studies were restricted to Latina women with variable nativity; in four studies, 26,54,56,57 Mexico was the country of or-igin for all or most of the women; and one additional study included a substantial percentage of Mexicanorigin Latinas. 55 The nine reviewed studies examined U.S. populations sourced from seven states; of these, two used data from three states traditionally receiving immigrants: Texas, California, and New York in one study 58 and California, New York, and Florida in another. 55 Four total studies utilized data from California, [54][55][56]58 three from New York, 47,55,58 two from Texas, 58,59 and one each for Utah, 26 Massachusetts, 60 Colorado, 57 and Florida 55 ( Table 1).
Documentation status was determined based on selfreport in three studies 54,55,60 ; the remaining studies relied on absence of social security number 26,47,56 and/or emergency Medicaid usage 57-59 as proxies for undocumented status. Notably, two of the three studies employing emergency Medicaid status as a proxy for undocumented status did not disclose the race/ethnicity 57 and/or country of origin. 59 Seven studies 26,47,[56][57][58][59][60] examined the impact of documentation status on birth outcomes. Each of these studies considered continuous birthweight (or dichotomized LBW), making it the most frequently assessed outcome. Three of the seven studies 26,57,59 examining birth outcomes found that undocumented status was associated with lower odds of dichotomized PTB and/or LBW infants ( Table 2). Two of these studies 26,57 specified this outcome among Mexican origin women; the remaining study 59 did not specify nativity but was based in Texas. In contrast, the authors of one study 26 observed that undocumented foreign-born Latinas had greater odds of giving birth to SGA infants than documented foreign-born Latinas before adjustment for maternal factors, including pregnancy complications; notably, this study yielded mixed results and also found no statistically significant associations with LBW and protective impacts on PTB. An additional study's 60 examination of continuous birthweight yielded a gradient whereby, on average, infants born to documented foreign-born mothers were the largest and infants born to U.S.-born mothers were the smallest.
Three studies 54,55,57 examined the relationship between documentation status and pregnancy outcomes. In minimally adjusted models, one 57 found that undocumented status was associated with higher odds of pregnancy complications and another 55 found that undocumented status was associated with postpartum depression. Adequate covariate adjustment was defined by our study team as adjustment for: maternal age, education, and marital status and was observed in none of the included studies. Two studies 54,60 included no   The nine studies overall met more than 60% of the quality parameters, with missing data being the most frequent study quality issue in this review. Notably, the proportion of the study sample with missing observations (and sociodemographic characteristics of those with missing observations) was seldom reported-all studies performed complete case analyses. In studies that did report on missingness, differential missingness was observed (e.g., 31% Children's Health Insurance Program [CHIP] prenatal vs. 10% Medicaid missing 59 ).
The number of quality parameters met by each study can be found in Appendix Tables A2 and A3, and the implications of unmet quality parameters are examined in the discussion.

Discussion
Based on our systematic review, evidence for the impact of documentation status on pregnancy outcomes among Latinas is not conclusive. Our hypothesisthat among foreign-born Latinas documented status would prove protective for pregnancy outcomes-was not wholly borne out, with our finding of divergent associations across outcomes. Undocumented status was generally either not associated or associated with lower odds of PTB and LBW; however, we also saw that being undocumented was associated with greater odds of pregnancy complications, abnormal conditions of the newborn, and postpartum depression. Given the heterogeneity of the studies (with regard to populations included, variable definitions of exposures and outcomes, and the diversity of contextual factors considered), the inconsistency was unsurprising. Until researchers engage in more standardized approaches, the true effects of documentation status on pregnancy outcomes may remain unclear.
One influence on our findings could be the different causal pathways leading to each unique outcome; elucidating these pathways has important implications for advancing health equity. For example, stress, which we hypothesized to be differentially experienced by documentation status and is a known risk factor for PTB and LBW, was found to be protective for these outcomes. This may be because this pathway is not as sensitive to immigration stress as expected, or that this stress is experienced too proximally to the pregnancy outcome to be adverse. Or it could be that all Latina women, whether documented or not, may be experiencing stress resulting from fear for family members or friends who may be undocumented and identified as such, or from having their own documentation status questioned. Therefore, the literature, as it stands, may be unable to distinguish the physiological stress resulting from documentation status from the chronic stress experienced by the Latina community overall. Given the evidence on social support among Latinas, it may also be that these strong relationships are, in fact, mitigating immigrant stress in ways that limit its adverse impacts, despite evidence that these ties are challenging for immigrants to maintain. Or it could be that the benefits associated with being foreign born are so strong that any impacts resulting from lack of documentation are not sufficiently adverse to neutralize them. We did expect to find that the protective effects conferred by foreign-born status would be diminished when compounded by undocumented status, with the lack of legal documentation ''overriding'' the protective effects resulting in the Latina paradox, and here our results and expectations aligned.
For those outcomes in our sample of studies not associated with physiological stress (e.g., pregnancy complications, postpartum depression, or unintended cesarean section), undocumented status was found to be associated with greater odds of occurrence. This set of outcomes is more directly related to poor patient-provider communication, inadequate prenatal care, or nonadherence to clinical recommendations and could, therefore, be more sensitive to documentation (and not protected by social ties), with undocumented status potentially leading to increased discomfort with or inability to communicate with medical care providers regarding pain experienced, birthing preferences, or other emerging issues. Future research examining the role of documentation status should ensure to consider the immigrant social ties hypothesis as well as patient-provider interactions to tease out these relationships. 61 In addition to these explanations, which focus on how documentation status is conceptualized to impact pregnancy outcomes, we identify multiple analytical issues that may have limited our ability to see clear relationships in the reviewed literature.

Challenge 1: population inclusivity
The most vulnerable undocumented women may not be properly represented in the reviewed literature. For studies conducting interviews with Latina women, differential representation may be based on challenges specific to the research questions (e.g., fears over revealing documentation status, linguistic barriers). In addition, in the absence of an interpreter, undocumented women may be systematically excluded for being disproportionately non-English speaking; one reviewed study 60 highlights this reliance on Spanish-language interpreter availability as an ultimate influence for participants. Further, undocumented Latinas may be underrepresented by virtue of where recruitment occurs, and even when included may be excluded from final models due to differential missingness. 59 Challenge 2: measurement of documentation status Revealed in this set of literature is a reliance on proxy measures for ascertaining documentation status. Our inconsistent findings may be due in part to this use of proxies since they provide an indirect assessment of the complex relationship between documentation status and pregnancy outcomes. For example, two studies 57,59 employed emergency Medicaid utilization as an indicator of undocumented status; however, this is apparently the closest approximation available, and it is merely a proxy for legal status. Three additional studies categorized participants as undocumented if they could not produce a social security number 26,56 or resident status card 47 at intake. Again, although a reasonable approach, this method does not guarantee specificity in measurement of documentation status and could result in miscategorization 49 and dilution of the potential effect of documentation status.

Challenge 3: heterogeneity of sampled population
Another potential cause of inconsistencies across the findings could stem from populations sampled: Of the nine reviewed studies, only five included predominantly Mexican foreign-born samples. The other studies included a range of Latina subgroups, which is important for our understanding of the findings because the Latina paradox is most robust among Mexican-born women and immigrant experience varies by nativity. 62 An example of this sampling issue is found in one study, 47 which included a sample that was only 14% Mexican-born, finding that docu-mentation status was not associated with LBW, a result that may be due to the authors' use of an all foreignborn comparison group, an analytic decision that could have resulted in a ''washing away'' of the effect of paradox. Across the reviewed literature, undocumented women were frequently compared with women of heterogeneous nativity status: Only three of the studies compared undocumented foreign-born Latinas with a group consisting solely of documented foreign-born Latinas or only U.S.-born Latinas. Because we hypothesized the poorest outcomes among our U.S.-born Latinas and the best outcomes among our documented foreign-born Latinas, null results could be explained by an even mixture of the two and potentially seemingly ''protective'' findings could be resulting from sampled populations where there are more U.S.-born than foreign-born Latinas.
Challenge 4: adjustment/model specification A number of studies included variables in their regression models that could actually be in the causal pathway between documentation status and pregnancy outcome. Specific covariates varying based on documentation status include: presence/number of prenatal care visits, insurance status, employment status, and substance use 56 ; smoking 58 ; income, employment, and health insurance coverage 55 ; and prepregnancy BMI, smoking, and alcohol use. 26 These factors are hypothesized to explain part of the relationship between documentation status and the observed outcomes, so this over-adjustment could bias results toward the null. In contrast, others 54,59,60 incorporate minimal or no covariate adjustment in their analyses and/or do not restrict evaluation of birth outcomes to singleton births.
Given mixed findings on the impact of documentation status on pregnancy outcomes across the reviewed literature, a full understanding of the relationship between this exposure and the outcomes of interest remains elusive. The outcomes reported here suggest that attempts to elucidate this relationship would be enhanced by clear theorizing on the pathways leading to impact; analytic strategies that reflect these conceptualizations; consistent measurement of documentation status and pregnancy outcomes; and sampling appropriate to investigating these relationships. We recommend that researchers clarify the specific ways in which they believe that being undocumented would impact pregnancy outcomes, which will guide the selection of appropriate outcomes to be examined. 50 In addition, we would hope to see this conceptual work reflected in the sample selection, which-if done in ways that recognize barriers to participation (e.g., legal status, language barriers)-would bolster findings and aid understanding of any identified associations. 50 Our final recommendation pertains to the measurement of documentation status: Researchers should limit the use of proxy measures when feasible. As previously discussed, the collection of legal status data carries tremendous risk for research participants, and public health researchers have been vocal about the need for caution in collecting this information. 50,52 The majority of our reviewed studies were published at least 20 years ago, and the climate around immigration and documentation status has shifted, leading toward an increasingly precarious position for undocumented research participants. Fortunately, guidance exists on how to engage in the collection of such data in ethical, scientifically responsible ways that can advance our understanding of documentation status impacts on health. Specific datasets (e.g., LA FANs) have incorporated precautions to protect undocumented participants and are considered well suited for such examinations. 48,51 A recent review of documentation status measurement in health research suggested a move away from proxy measures to self-report, 49 building on prior recommendations to combine survey and ethnographic approaches. 63 Others 50 offer specific strategies for protecting participants while ensuring data validity, and addressing issues related to navigating IRB applications and securing informed consent, providing 51 a model for how to engage in this work in culturally responsive ways. Ultimately, legal protections and controls are warranted when asking for this information. 52 Our review has some important limitations. First, our search was limited to articles that were written in English, which may introduce language bias. However, because the focus of our study was the United States, we believe that there are less likely to be missing papers written in other languages. Second, although we performed a robust review including Google Scholarwhich is useful for finding gray literature-our review may be subject to publication bias. Third, because of the heterogeneity in outcomes, exposure definitions, and groups being compared, we were unable to conduct a formal meta-analysis. Fourth, despite including several key words for pregnancy outcomes that occur disproportionately among Latina women, such as GWG, gestational diabetes, and LGA, none of the eligible studies examined these outcomes. Finally, the small sample size coupled with afore mentioned heterogeneity reduced our ability to make strong inferences about the relationship between documentation status and pregnancy and birth outcomes in Latina women.
To our knowledge, this work constitutes the first systematic review of the impacts of documentation status on Latina pregnancy outcomes. Our findings highlight the need for further examination of the role of legal status by showing that being undocumented in the United States can adversely impact the health of women and their offspring, with far-reaching potential for health and health inequities across the life course. Researchers engaging in this work should consider the challenges we describe here-related to theory, sampling, measurement, and modeling-and consider the related recommendations when developing studies to examine documentation status and pregnancy outcomes among Latinas.

Q51 What Type of Study Design?
, Ecological (1) , Cross-Sectional (2) , Prospective Cohort (6) , A paragraph in the methods section of the article may explain the sample size needed to detect a hypothesized difference in outcomes. You may also find a discussion of power in the discussion section (such as the study had 85% power to detect a 20% increase in the rate of an outcome of interest, with a two-sided alpha of 0.05). Sometimes, estimates of variance and/or estimates of effect size are given, instead of sample size calculations.
In any of these cases, the answer would be ''yes.'' However, observational cohort studies often do not report anything about power or sample sizes because the analyses are exploratory in nature. In this case, the answer would be ''no.'' This is not a ''fatal flaw.'' It just may indicate that attention was not paid to whether the study was sufficiently sized to answer a prespecified question-that is, it may have been an exploratory, hypothesisgenerating study. If a cohort study is conducted properly, the answer to this question should be ''yes,'' since the exposure status of members of the cohort was determined at the beginning of the study before the outcomes occurred. Because in retrospective cohort studies the exposure and outcomes may have already occurred (it depends on how long they follow the cohort), it is important to make sure that the exposure preceded the outcome. Sometimes, cross-sectional studies are conducted (or cross-sectional analyses of cohort-study data), where the exposures and outcomes are measured during the same timeframe. As a result, cross-sectional analyses provide weaker evidence than regular cohort studies regarding a potential causal relationship between exposures and outcomes. For cross-sectional analyses, the answer to Question 6 should be ''no.'' B Yes (1)  Blinding means that outcome assessors did not know whether the participant was exposed or unexposed. It is also sometimes called ''masking.'' The objective is to look for evidence in the article that the person(s) assessing the outcome(s) for the study (for example, examining medical records to determine the outcomes that occurred in the exposed and comparison groups) is masked to the exposure status of the participant. Sometimes, the person measuring the exposure is the same person conducting the outcome assessment. In this case, the outcome assessor would most likely not be blinded to exposure status because they also took measurements of exposures. If so, make a note of that in the comments section. As you assess this criterion, think about whether it is likely that the person(s) doing the outcome assessment would know (or be able to figure out) the exposure status of the study participants. If the answer is no, then blinding is adequate. An example of adequate blinding of the outcome assessors is to create a separate committee, whose members were not involved in the care of the patient and had no information about the study participants' exposure status. The committee would then be provided with copies of participants' medical records, which had been stripped of any potential exposure information or personally identifiable information. The committee would then review the records for prespecified outcomes according to the study protocol. If blinding was not possible, which is sometimes the case, mark ''NA'' and explain the potential for bias. Higher overall follow-up rates are always better than lower follow-up rates, even though higher rates are expected in shorter studies, whereas lower overall follow-up rates are often seen in studies of longer duration. Usually, an acceptable overall follow-up rate is considered 80% or more of participants whose exposures were measured at baseline. However, this is just a general guideline. For example, a 6-month cohort study examining the relationship between dietary sodium intake and BP level may have more than 90% follow-up, but a 20-year cohort study examining effects of sodium intake on stroke may have only a 65% follow-up rate. Were key potential confounding variables measured and adjusted statistically for their impact on the relationship between exposure(s) and outcome(s)?
Were key potential confounding variables measured and adjusted for, such as by statistical adjustment for baseline differences? Logistic regression or other regression methods are often used to account for the influence of variables that are not of interest. This is a key issue in cohort studies, because statistical analyses need to control for potential confounders, in contrast to an RCT, where the randomization process controls for potential confounders. All key factors that may be associated with both the exposure of interest and the outcome-which are not of interest to the research question-should be controlled for in the analyses. The questions just cited are designed to help you focus on the key concepts for evaluating the internal validity of a study. They are not intended to create a list that you simply tally to arrive at a summary judgment of quality.
Internal validity for cohort studies is the extent to which the results reported in the study can truly be attributed to the exposure being evaluated and not to flaws in the design or conduct of the study-in other words, the ability of the study to draw associative conclusions about the effects of the exposures being studied on outcomes. Any such flaws can increase the risk of bias. Critical appraisal involves considering the risk of potential for selection bias, information bias, measurement bias, or confounding (the mixture of exposures that one cannot tease out from each other). Examples of confounding include co-interventions, differences at baseline in patient characteristics, and other issues throughout the questions cited earlier. High risk of bias translates to a rating of poor quality. Low risk of bias translates to a rating of good quality. (Thus, the greater the risk of bias, the lower the quality rating of the study.) In addition, the more attention in the study design to issues that can help determine whether there is a causal relationship between the exposure and outcome, the higher the quality of the study. These include exposures occurring before outcomes, evaluation of a dose-response gradient, accuracy of measurement of both exposure and outcome, sufficient timeframe to see an effect, and appropriate control for confounding-all concepts reflected in the tool. Generally, when you evaluate a study, you will not see a ''fatal flaw,'' but you will find some risk of bias. By focusing on the concepts underlying the questions in the quality assessment tool, you should ask yourself about the potential for bias in the study you are critically appraising. For any box where you check ''no,'' you should ask, ''What is the potential risk of bias resulting from this flaw in study design or execution?'' That is, does this factor cause you to doubt the results that are reported in the study or doubt the ability of the study to accurately assess an association between exposure and outcome?
The best approach is to think about the questions in the tool and how each one tells you something about the potential for bias in a study. The more you familiarize yourself with the key concepts, the more comfortable you will be with critical appraisal. Examples of studies rated good, fair, and poor are useful, but each study must be assessed on its own based on the details that are reported and consideration of the concepts for minimizing bias. Q67 Final thoughts on quality review to discuss among the group: ___________________________________________________________________________________ ___________________________________________________________________________________ ___________________________________________________________________________________ ___________________________________________________________________________________ ___________________________________________________________________________________ Q40 Comments or additional information not previously indicated: ___________________________________________________________________________________ ___________________________________________________________________________________ ___________________________________________________________________________________ ___________________________________________________________________________________