It's not all about residual confounding: a plea for QBA for epidemiologic researchers and educators

We spend a great deal of time on confounding in our teaching, in our methods development and in our assessment of study results. This may give the impression that uncontrolled confounding is the biggest problem that observational epidemiology faces, when in fact, other sources of bias such as selection bias, measurement error, missing data, and misalignment of zero time may often (especially if they are all present in a single study) lead to a stronger deviation from the truth. Compared to the amount of time we spend teaching how to address confounding in a data analysis, we spend relatively little time teaching methods for simulating confounding (and other sources of bias) to learn their impact and develop plans to mitigate or quantify the bias. We review a paper by Desai et al that uses simulation methods to quantify the impact of an unmeasured confounder when it is completely missing or when a proxy of the confounder is measured. We use this article to discuss how we can use simulations of sources of bias to ensure we generate better and more valid study estimates, and we discuss the importance of simulating realistic datasets with plausible bias structures to guide data collection. If an advanced life form exists outside of our current universe and they came to earth with the goal of scouring the published epidemiologic literature to understand what the biggest problem epidemiologists have, they would quickly discover that the limitations section of publications would provide them with all the information they needed. And most likely what they would conclude is that the biggest problem that we face is uncontrolled confounding. It seems to be an obsession of ours.

Use of quantitative bias analysis to evaluate single-arm trials with real-world data external controls

PURPOSE

Use of real-world data (RWD) for external controls added to single-arm trials (SAT) is increasingly prevalent in regulatory submissions. Due to inherent differences in the data-generating mechanisms, biases can arise. This paper aims to illustrate how to use quantitative bias analysis (QBA).

METHODS

Advanced non-small cell lung cancer (NSCLC) serves as an example, where many small subsets of patients with molecular tumor subtypes exist. First, some sources of bias that may occur in oncology when comparing RWD to SAT are described. Second, using a hypothetical immunotherapy agent, a dataset is simulated based on expert input for survival analysis of advanced NSCLC. Finally, we illustrate the impact of three biases: missing confounder, misclassification of exposure, and outcome evaluation.

RESULTS

For each simulated scenario, bias was induced by removing or adding data; hazard ratios (HRs) were estimated applying conventional analyses. Estimating the bias-adjusted treatment effect and uncertainty required carefully selecting the bias model and bias factors. Although the magnitude of each biased and bias-adjusted HR appeared moderate in all three hypothetical scenarios, the direction of bias was variable.

CONCLUSION

These findings suggest that QBA can provide an intuitive framework for bias analysis, providing a key means of challenging assumptions about the evidence. However, the accuracy of bias analysis is itself dependent on correct specification of the bias model and bias factors. Ultimately, study design should reduce bias, but QBA allows us to evaluate the impact of unavoidable bias to assess the quality of the evidence.

Paediatric-onset immune-mediated inflammatory disease is associated with an increased mortality risk-A nationwide study

BACKGROUND

Paediatric-onset immune-mediated inflammatory diseases (pIMID) show more aggressive phenotypes than when diagnosed in adults. However, data on mortality are often extrapolated from adult studies.

AIM

To estimate the effect of pIMID on mortality.

METHODS

In a population-based cohort study using the nationwide Danish healthcare registers, we included all patients diagnosed with pIMID in Denmark from 1980 to 2018. PIMID were defined as ICD codes indicative of autoimmune hepatitis, primary sclerosing cholangitis, Crohn's disease, ulcerative colitis, juvenile idiopathic arthritis, lupus erythematosus, or vasculitis registered before age 18 years. All-cause mortality was the primary outcome; cause-specific mortality was the secondary outcome. We used Cox survival analysis to estimate hazard ratios (HR), and Aalen survival analysis to estimate rate differences.

RESULTS

We included 11,581 individuals diagnosed with pIMID and 99,665 reference individuals, accounting for 1,371,994 person-years of follow-up. Median and interquartile (IQR) age at diagnosis was 12.6 (7.9-15.9) years. During follow-up, 152 patients with pIMID and 316 reference individuals died; adjusted HR (aHR) was 3.8 (95% confidence interval [CI] 3.1-4.7). This corresponded to 6.9 (95% CI: 5.3-8.5) additional deaths per 10,000 person-years. The strongest associations were found for gastrointestinal diseases (aHR 22.8; 95% CI 9.6-64.1), gastrointestinal cancers (aHR 19.2; 95% CI 5.0-74.2) and lymphoproliferative disorders (aHR 6.8; 95% CI 2.8-16.8).

CONCLUSION

Patients diagnosed with pIMID have a fourfold higher risk of mortality when followed into early adulthood compared with reference individuals. This underlines the severe disease course of pIMID and highlights the need for multidisciplinary care.

Ambulatory Extracorporeal Membrane Oxygenation Use in Patients Listed for Lung Transplant

There has been an increase in the use of extracorporeal membrane oxygenation (ECMO) to bridge critically ill patients to lung transplant (LTX). This study evaluates how ambulatory status on ECMO affected waitlist and post-LTX outcomes. The United Network of Organ Sharing (UNOS) database was queried for patients aged of greater than or equal to 18 years and between 2016 and 2021 to identify pre-LTX patients supported by ECMO. The patients were classified in venous-arterial (VA) ECMO and veno-venous (VV) ECMO cohorts and further classified as ambulatory (AMB) and non-AMB (nAMB). Each cohort was controlled against the non-ECMO patients. Univariate statistical tests, as well as Kaplan-Meier survival curves, were used for analysis. The 90 day waitlist survival was the highest among the non-ECMO group (96%), but both AMB VV and VA groups had superior survival compared to the nAMB group (85% vs. 75%, 78% vs. 65%, p < 0.01). After adjusting for the median lung allocation score (LAS) (88) in the VV ECMO group, the waitlist survival was superior in the AMB VV ECMO compared to those not on ECMO (86% vs. 78%, p > 0.01). The 1 year post-LTX survival between non-ECMO and AMB VV ECMO was comparable (88% vs. 88%, p = 0.66). Ambulating patients or use of physical therapy while on ECMO can help improve lung transplant outcomes.

Does Engagement in HIV Care Affect Screening, Diagnosis, and Control of Noncommunicable Diseases in Sub-Saharan Africa? A Systematic Review and Meta-analysis

Low- and middle-income countries are facing a growing burden of noncommunicable diseases (NCDs). Providing HIV treatment may provide opportunities to increase access to NCD services in under-resourced environments. We conducted a systematic review and meta-analysis to evaluate whether use of antiretroviral therapy (ART) was associated with increased screening, diagnosis, treatment, and control of diabetes, hypertension, chronic kidney disease, or cardiovascular disease among people living with HIV in sub-Saharan Africa (SSA). A comprehensive search of electronic literature databases for studies published between 01 January 2011 and 31 December 2022 yielded 26 studies, describing 13,570 PLWH in SSA, 61% of whom were receiving ART. Random effects models were used to calculate summary odds ratios (ORs) of the risk of diagnosis by ART status and corresponding 95% confidence intervals (95% CIs), where appropriate. ART use was associated with a small but imprecise increase in the odds of diabetes diagnosis (OR 1.07; 95% CI 0.71, 1.60) and an increase in the odds of hypertension diagnosis (OR 2.10, 95% CI 1.42, 3.09). We found minimal data on the association between ART use and screening, treatment, or control of NCDs. Despite a potentially higher NCD risk among PLWH and regional efforts to integrate NCD and HIV care, evidence to support effective care integration models is lacking.

Gaps in the type 2 diabetes care cascade: a national perspective using South Africa's National Health Laboratory Service (NHLS) database

BACKGROUND

Research out of South Africa estimates the total unmet need for care for those with type 2 diabetes mellitus (diabetes) at 80%. We evaluated the care cascade using South Africa's National Health Laboratory Service (NHLS) database and assessed if HIV infection impacts progression through its stages.

METHODS

The cohort includes patients from government facilities with their first glycated hemoglobin A1c (HbA1c) or plasma glucose (fasting (FPG); random (RPG)) measured between January 2012 to March 2015 in the NHLS. Lab-diagnosed diabetes was defined as HbA1c ≥ 6.5%, FPG ≥ 7.0mmol/l, or RPG ≥ 11.1mmol/l. Cascade stages post diagnosis were retention-in-care and glycaemic control (defined as an HbA1c < 7.0% or FPG < 8.0mmol/l or RPG < 10.0mmol/l) over 24-months. We estimated gaps at each stage nationally and by people living with HIV (PLWH) and without (PLWOH).

RESULTS

Of the 373,889 patients tested for diabetes, 43.2% had an HbA1c or blood glucose measure indicating a diabetes diagnosis. Amongst those with lab-diagnosed diabetes, 30.9% were retained-in-care (based on diabetes labs) and 8.7% reached glycaemic control by 24-months. Prevalence of lab-diagnosed diabetes in PLWH was 28.6% versus 47.3% in PLWOH. Among those with lab-diagnosed diabetes, 34.3% of PLWH were retained-in-care versus 30.3% PLWOH. Among people retained-in-care, 33.8% of PLWH reached glycaemic control over 24-months versus 28.6% of PLWOH.

CONCLUSIONS

In our analysis of South Africa's NHLS database, we observed that 70% of patients diagnosed with diabetes did not maintain in consistent diabetes care, with fewer than 10% reaching glycemic control within 24 months. We noted a disparity in diabetes prevalence between PLWH and PLWOH, potentially linked to different screening methods. These differences underscore the intricacies in care but also emphasize how HIV care practices could guide better management of chronic diseases like diabetes. Our results underscore the imperative for specialized strategies to bolster diabetes care in South Africa.

Associations of inter-annual rainfall decreases with subsequent HIV outcomes for persons with HIV on antiretroviral therapy in Southern Africa: a collaborative analysis of cohort studies

BACKGROUND

Periods of droughts can lead to decreased food security, and altered behaviours, potentially affecting outcomes on antiretroviral therapy (ART) among persons with HIV (PWH). We investigated whether decreased rainfall is associated with adverse outcomes among PWH on ART in Southern Africa.

METHODS

Data were combined from 11 clinical cohorts of PWH in Lesotho, Malawi, Mozambique, South Africa, Zambia, and Zimbabwe, participating in the International epidemiology Databases to Evaluate AIDS Southern Africa (IeDEA-SA) collaboration. Adult PWH who had started ART prior to 01/06/2016 and were in follow-up in the year prior to 01/06/2016 were included. Two-year rainfall from June 2014 to May 2016 at the location of each HIV centre was summed and ranked against historical 2-year rainfall amounts (1981-2016) to give an empirical relative percentile rainfall estimate. The IeDEA-SA and rainfall data were combined using each HIV centre's latitude/longitude. In individual-level analyses, multivariable Cox or generalized estimating equation regression models (GEEs) assessed associations between decreased rainfall versus historical levels and four separate outcomes (mortality, CD4 counts < 200 cells/mm3, viral loads > 400 copies/mL, and > 12-month gaps in follow-up) in the two years following the rainfall period. GEEs were used to investigate the association between relative rainfall and monthly numbers of unique visitors per HIV centre.

RESULTS

Among 270,708 PWH across 386 HIV centres (67% female, median age 39 [IQR: 32-46]), lower rainfall than usual was associated with higher mortality (adjusted Hazard Ratio: 1.18 [95%CI: 1.07-1.32] per 10 percentile rainfall rank decrease) and unsuppressed viral loads (adjusted Odds Ratio: 1.05 [1.01-1.09]). Levels of rainfall were not strongly associated with CD4 counts < 200 cell/mm3 or > 12-month gaps in care. HIV centres in areas with less rainfall than usual had lower numbers of PWH visiting them (adjusted Rate Ratio: 0.80 [0.66-0.98] per 10 percentile rainfall rank decrease).

CONCLUSIONS

Decreased rainfall could negatively impact on HIV treatment behaviours and outcomes. Further research is needed to explore the reasons for these effects. Interventions to mitigate the health impact of severe weather events are required.

Moderate alcohol consumption on the risk of stroke in the Million Veteran Program

BACKGROUND

There is inconsistent evidence on the association of moderate alcohol consumption and stroke risk in the general population and is not well studied among U.S. Veterans. Furthermore, it is unclear whether primarily drinking beer, wine, or liquor is associated with a difference in stroke risk.

METHODS

The study included 185,323 Million Veteran Program participants who self-reported alcohol consumption on the Lifestyle Survey. Moderate consumption was defined as 1-2 drinks/day and beverage preference of beer, wine or liquor was defined if ≥ 50% of total drinks consumed were from a single type of beverage. Strokes were defined using ICD-9 and ICD-10 codes from the participants' electronic health record.

RESULTS

The mean (sd) age of the sample was 64 (13) years and 11% were women. We observed 4,339 (94% ischemic; 6% hemorrhagic) strokes over a median follow-up of 5.2 years. In Cox models adjusted for age, sex, race, education, income, body mass index, smoking, exercise, diet, cholesterol, prevalent diabetes, prevalent hypertension, lipid-lowering medication, antihypertensive medication, and diabetes medication, moderate alcohol consumption (1-2 drinks/day) was associated with a 22% lower risk of total stroke compared with never drinking [Hazards ratio (HR) 95% confidence interval (CI): 0.78 (0.67, 0.92)]. When stratifying by stroke type, we observed a similar protective association with moderate consumption and ischemic stroke [HR (95% CI): 0.76 (0.65, 0.90)], but a non-statistically significant higher risk of hemorrhagic stroke [HR (95% CI): 1.29 (0.64, 2.61)]. We did not observe a difference in ischemic or hemorrhagic stroke risk among those who preferred beer, liquor or wine vs. no beverage preference. When stratifying by prior number of hospital visits (≤ 15, 16-33, 34-64, ≥ 65) as a proxy for health status, we observed attenuation of the protective association with greater number of visits [HR (95% CI): 0.87 (0.63, 1.19) for ≥ 65 visits vs. 0.80 (0.59, 1.08) for ≤ 15 visits].

CONCLUSIONS

We observed a lower risk of ischemic stroke, but not hemorrhagic stroke with moderate alcohol consumption and did not observe substantial differences in risk by beverage preference among a sample of U.S. Veterans. Healthy user bias of moderate alcohol consumption may be driving some of the observed protective association.

Quantitative Assessment of Systematic Bias: A Guide for Researchers

Observational research provides valuable opportunities to advance oral health science but is limited by vulnerabilities to systematic bias, including unmeasured confounding, errors in variable measurement, or bias in the creation of study populations and/or analytic samples. The potential influence of systematic biases on observed results is often only briefly mentioned among the discussion of limitations of a given study, despite existing methods that support detailed assessments of their potential effects. Quantitative bias analysis is a set of methodological techniques that, when applied to observational data, can provide important context to aid in the interpretation and integration of observational research findings into the broader body of oral health research. Specifically, these methods were developed to provide quantitative estimates of the potential magnitude and direction of the influence of systematic biases on observed results. We aim to encourage and facilitate the broad adoption of quantitative bias analyses into observational oral health research. To this end, we provide an overview of quantitative bias analysis techniques, including a step-by-step implementation guide. We also provide a detailed appendix that guides readers through an applied example using real data obtained from a prospective observational cohort study of preconception periodontitis in relation to time to pregnancy. Quantitative bias analysis methods are available to all investigators. When appropriately applied to observational studies, findings from such studies can have a greater impact in the broader research context.

SAS and R code for probabilistic quantitative bias analysis for misclassified binary variables and binary unmeasured confounders

Systematic error from selection bias, uncontrolled confounding, and misclassification is ubiquitous in epidemiologic research but is rarely quantified using quantitative bias analysis (QBA). This gap may in part be due to the lack of readily modifiable software to implement these methods. Our objective is to provide computing code that can be tailored to an analyst's dataset. We briefly describe the methods for implementing QBA for misclassification and uncontrolled confounding and present the reader with example code for how such bias analyses, using both summary-level data and individual record-level data, can be implemented in both SAS and R. Our examples show how adjustment for uncontrolled confounding and misclassification can be implemented. Resulting bias-adjusted point estimates can then be compared to conventional results to see the impact of this bias in terms of its direction and magnitude. Further, we show how 95% simulation intervals can be generated that can be compared to conventional 95% confidence intervals to see the impact of the bias on uncertainty. Having easy to implement code that users can apply to their own datasets will hopefully help spur more frequent use of these methods and prevent poor inferences drawn from studies that do not quantify the impact of systematic error on their results.

Health systems performance for hypertension control using a cascade of care approach in South Africa, 2011-2017

Hypertension is a major contributor to global morbidity and mortality. In South Africa, the government has employed a whole systems approach to address the growing burden of non-communicable diseases. We used a novel incident care cascade approach to measure changes in the South African health system's ability to manage hypertension between 2011 and 2017. We used data from Waves 1-5 of the National Income Dynamics Study (NIDS) to estimate trends in the hypertension care cascade and unmet treatment need across four successive cohorts with incident hypertension. We used a negative binomial regression to identify factors that may predict higher rates of hypertension control, controlling for socio-demographic and healthcare factors. In 2011, 19.6% (95%CI 14.2, 26.2) of individuals with incident hypertension were diagnosed, 15.4% (95%CI 10.8, 21.4) were on treatment and 7.1% had controlled blood pressure. By 2017, the proportion of individuals with diagnosed incident hypertension had increased to 24.4% (95%CI 15.9, 35.4). Increases in treatment (23.3%, 95%CI 15.0, 34.3) and control (22.1%, 95%CI 14.1, 33.0) were also observed, translating to a decrease in unmet need for hypertension care from 92.9% in 2011 to 77.9% in 2017. Multivariable regression showed that participants with incident hypertension in 2017 were 3.01 (95%CI 1.77, 5.13) times more likely to have a controlled blood pressure compared to those in 2011. Our data show that while substantial improvements in the hypertension care cascade occurred between 2011 and 2017, a large burden of unmet need remains. The greatest losses in the incident hypertension care cascades came before diagnosis. Nevertheless, whole system programming will be needed to sufficiently address significant morbidity and mortality related to having an elevated blood pressure.

Preconception Periodontitis and Risk of Spontaneous Abortion in a Prospective Cohort Study

Few studies have evaluated the association between periodontitis and spontaneous abortion (SAB), and all had limitations. We used data from the Pregnancy Study Online (PRESTO), a prospective preconception cohort study of 3,444 pregnancy planners in the United States and Canada (2019-2022), to address this question. Participants provided self-reported data on periodontitis diagnosis, treatment, and symptoms of severity (i.e., loose teeth) via the enrollment questionnaire. SAB (pregnancy loss at <20 weeks' gestation) was assessed via bimonthly follow-up questionnaires. Participants contributed person-time from the date of a positive pregnancy test to the gestational week of SAB, loss to follow-up, or 20 weeks' gestation, whichever came first. We fitted Cox regression models with weeks of gestation as the time scale to estimate adjusted hazard ratios (HRs) and 95% confidence intervals (CIs), and we used inverse probability of treatment weighting to account for differential loss to follow-up. We used probabilistic quantitative bias analysis to estimate the magnitude and direction of the effect of exposure misclassification bias on results. In weighted multivariable models, we saw no appreciable association between preconception periodontitis diagnosis (HR = 0.97, 95% CI: 0.76, 1.23) or treatment (HR = 1.01, 95% CI: 0.79, 1.27) and SAB. A history of loose teeth was positively associated with SAB (HR = 1.38, 95% CI: 0.88, 2.14). Quantitative bias analysis indicated that our findings were biased towards the null but with considerable uncertainty in the bias-adjusted results.

The Association Between Immune-Related Conditions Across the Life-Course and Provoked Vulvodynia

Vulvodynia, impacts up to 8% of women by age 40, and is hypothesized to manifest through an altered immune-inflammatory response. To test this hypothesis, we identified all women born in Sweden between 1973 and 1996 diagnosed with localized provoked vulvodynia (N76.3) and/or vaginismus (N94.2 or F52.5) between 2001 and 2018. We matched each case to two women from the same birth year with no vulvar pain ICD codes. As a proxy for immune dysfunction, we used Swedish Registry data to capture 1) immunodeficiencies, 2) single organ and multiorgan autoimmune conditions, 3) allergy and atopies, and 4) malignancies involving immune cells across the life course. Women with vulvodynia, vaginismus or both were more likely to experience immune deficiencies (OR 1.8, 95% CI, 1.2-2.8), single organ (OR 1.4, 95% CI, 1.2-1.6) and/or multi-organ (OR 1.6, 95% CI, 1.3-1.9) immune disorders, and allergy/atopy conditions (OR 1.7, 95% CI, 1.6-1.8) compared to controls. We observed greater risk with increasing numbers of unique immune related conditions (1 code: OR = 1.6, 95% CI, 1.5-1.7; 2 codes: OR = 2.4, 95% CI, 2.1-2.9; 3 or more codes: OR = 2.9, 1.6-5.4). These findings suggest that women with vulvodynia may have a more compromised immune system either at birth or at points across the life course than women with no vulvar pain history. PERSPECTIVE: Women with vulvodynia are substantially more likely to experience a spectrum of immune related conditions across the life course. These findings lend support to the hypothesis that chronic inflammation initiates the hyperinnervation that causes the debilitating pain in women with vulvodynia.

Historical visit attendance as predictor of treatment interruption in South African HIV patients: Extension of a validated machine learning model

Retention of antiretroviral (ART) patients is a priority for achieving HIV epidemic control in South Africa. While machine-learning methods are being increasingly utilised to identify high risk populations for suboptimal HIV service utilisation, they are limited in terms of explaining relationships between predictors. To further understand these relationships, we implemented machine learning methods optimised for predictive power and traditional statistical methods. We used routinely collected electronic medical record (EMR) data to evaluate longitudinal predictors of lost-to-follow up (LTFU) and temporal interruptions in treatment (IIT) in the first two years of treatment for ART patients in the Gauteng and North West provinces of South Africa. Of the 191,162 ART patients and 1,833,248 visits analysed, 49% experienced at least one IIT and 85% of those returned for a subsequent clinical visit. Patients iteratively transition in and out of treatment indicating that ART retention in South Africa is likely underestimated. Historical visit attendance is shown to be predictive of IIT using machine learning, log binomial regression and survival analyses. Using a previously developed categorical boosting (CatBoost) algorithm, we demonstrate that historical visit attendance alone is able to predict almost half of next missed visits. With the addition of baseline demographic and clinical features, this model is able to predict up to 60% of next missed ART visits with a sensitivity of 61.9% (95% CI: 61.5-62.3%), specificity of 66.5% (95% CI: 66.4-66.7%), and positive predictive value of 19.7% (95% CI: 19.5-19.9%). While the full usage of this model is relevant for settings where infrastructure exists to extract EMR data and run computations in real-time, historical visits attendance alone can be used to identify those at risk of disengaging from HIV care in the absence of other behavioural or observable risk factors.

Does engagement in HIV care affect screening, diagnosis, and control of noncommunicable diseases in sub-Saharan Africa? A systematic review and meta-analysis

Objective

Low- and middle-income countries are facing a growing burden of noncommunicable diseases (NCDs). Providing HIV treatment may also provide opportunities to increase access to NCD services in under-resourced environments. We sought to investigate whether reported use of antiretroviral therapy (ART) was associated with increased screening, diagnosis, treatment, and/or control of diabetes, hypertension, chronic kidney disease, or cardiovascular disease among people living with HIV (PLWH) in sub-Saharan Africa (SSA).

Design

Systematic review and meta-analysis.

Methods

We searched 10 electronic literature databases for studies published between 01 January 2011 and 31 December 2022 using a comprehensive search strategy. We sought studies reporting on screening, diagnosis, treatment, and/or control of NCDs of interest by ART use among non-pregnant adults with HIV ≥16 years of age in SSA. Random effects models were used to calculate summary odds ratios (ORs) of the risk of diagnosis by ART status and corresponding 95% confidence intervals (95% CIs), where appropriate.

Results

Twenty-six studies, describing 13,570 PLWH in SSA, 61% of whom were receiving ART, were included. ART use was associated with a small but imprecise increase in the odds of diabetes diagnosis (OR: 1.07; 95% CI: 0.71, 1.60) and an increase in the odds of hypertension diagnosis (OR: 2.10, 95% CI: 1.42, 3.09). We found minimal data on the association between ART use and screening, treatment, or control of NCDs.

Conclusion

Despite a potentially higher NCD risk among PLWH and regional efforts to integrate NCD and HIV care, evidence to support effective care integration models is lacking.

Burden of Chronic Obstructive Pulmonary Disease Attributable to Tuberculosis: A Microsimulation Study

Tuberculosis (TB) is a risk factor for chronic obstructive pulmonary disease (COPD), but COPD is also a predictor of TB. The excess life-years lost to COPD caused by TB can potentially be saved by screening for and treating TB infection. We examined the number of life-years that could be saved by preventing TB and TB-attributable COPD. We compared the observed (no intervention) and counterfactual microsimulation models constructed from observed rates in the Danish National Patient Registry (covering all Danish hospitals between 1995 and 2014). In the Danish population of TB and COPD-naive individuals (n = 5,206,922), 27,783 persons (0.5%) developed TB. Among those who developed TB, 14,438 (52.0%) developed TB with COPD. Preventing TB saved 186,469 life-years overall. The excess number of life-years lost to TB alone was 7.07 years per person, and the additional number of life-years lost among persons who developed COPD after TB was 4.86 years per person. The life-years lost to TB-associated COPD are substantial, even in regions where TB can be expected to be identified and treated promptly. Prevention of TB could prevent a substantial amount of COPD-related morbidity; the benefit of screening and treatment for TB infection is underestimated by considering morbidity from TB alone.

First-line opioid agonist treatment as prevention against assisting others in initiating injection drug use: A longitudinal cohort study of people who inject drugs in Vancouver, Canada

Background

Among people who inject drugs, frequent injecting and experiencing withdrawal are associated with facilitating others' first injections. As these factors may reflect an underlying substance use disorder, we investigated whether first-line oral opioid agonist treatment (OAT; methadone or buprenorphine/naloxone) reduces the likelihood that people who inject drugs help others initiate injecting.

Methods

We used questionnaire data from semi-annual visits between December 2014-May 2018 on 334 people who inject drugs with frequent non-medical opioid use in Vancouver, Canada. We estimated the effect of current first-line OAT on subsequent injection initiation assistance provision (i.e., helped someone initiate injecting in the following six months) using inverse-probability-weighted estimation of repeated measures marginal structural models to reduce confounding and informative censoring by time-fixed and time-varying covariates.

Results

By follow-up visit, 54-64% of participants reported current first-line OAT whereas 3.4-6.9% provided subsequent injection initiation assistance. Per the primary weighted estimate (n = 1114 person-visits), participants currently on first-line OAT (versus no OAT) were 50% less likely, on average, to subsequently help someone initiate injecting (relative risk [RR]=0.50, 95% CI=0.23-1.11). First-line OAT was associated with reduced risk of subsequent injection initiation assistance provision in participants who, at baseline, injected opioids less than daily (RR=0.15, 95% CI=0.05-0.44) but not in those who injected opioids daily (RR=0.86, 95% CI=0.35-2.11).

Conclusions

First-line OAT seemingly reduces the short-term likelihood that people who inject drugs facilitate first injections. However, the extent of this potential effect remains uncertain due to imprecise estimation and observed heterogeneity by baseline opioid injecting frequency.

Record linkage without patient identifiers: proof of concept using data from South Africa's national HIV program

Background

Linkage between health databases typically requires identifiers such as patient names and personal identification numbers. We developed and validated a record linkage strategy to combine administrative health databases without the use of patient identifiers, with application to South Africa's public sector HIV treatment program.

Methods

We linked CD4 counts and HIV viral loads from South Africa's HIV clinical monitoring database (TIER.Net) and the National Health Laboratory Service (NHLS) for patients receiving care between 2015-2019 in Ekurhuleni District (Gauteng Province). We used a combination of variables related to lab results contained in both databases (result value; specimen collection date; facility of collection; patient year and month of birth; and sex). Exact matching linked on exact linking variable values while caliper matching applied exact matching with linkage on approximate test dates (± 5 days). We then developed a sequential linkage approach utilising specimen barcode matching, then exact matching, and lastly caliper matching. Performance measures were sensitivity and positive predictive value (PPV); share of patients linked across databases; and percent increase in data points for each linkage approach.

Results

We attempted to link 2,017,290 lab results from TIER.Net (representing 523,558 unique patients) and 2,414,059 lab results from the NHLS database. Linkage performance was evaluated using specimen barcodes (available for a minority of records in TIER.net) as a "gold standard". Exact matching achieved a sensitivity of 69.0% and PPV of 95.1%. Caliper-matching achieved a sensitivity of 75.7% and PPV of 94.5%. In sequential linkage, we matched 41.9% of TIER.Net labs by specimen barcodes, 51.3% by exact matching, and 6.8% by caliper matching, for a total of 71.9% of labs matched, with PPV=96.8% and Sensitivity = 85.9%. The sequential approach linked 86.0% of TIER.Net patients with at least one lab result to the NHLS database (N=1,450,087). Linkage to the NHLS Cohort increased the number of laboratory results associated with TIER.Net patients by 62.6%.

Conclusions

Linkage of TIER.Net and NHLS without patient identifiers attained high accuracy and yield without compromising patient privacy. The integrated cohort provides a more complete view of patients' lab history and could yield more accurate estimates of HIV program indicators.

Toward Open and Reproducible Epidemiology

Starting in the 2010s, researchers in the experimental social sciences rapidly began to adopt increasingly open and reproducible scientific practices. These practices include publicly sharing deidentified data when possible, sharing analytical code, and preregistering study protocols. Empirical evidence from the social sciences suggests such practices are feasible, can improve analytical reproducibility, and can reduce selective reporting. In academic epidemiology, adoption of open-science practices has been slower than in the social sciences (with some notable exceptions, such as registering clinical trials). Epidemiologic studies are often large, complex, conceived after data have already been collected, and difficult to replicate directly by collecting new data. These characteristics make it especially important to ensure their integrity and analytical reproducibility. Open-science practices can also pay immediate dividends to researchers' own work by clarifying scientific reasoning and encouraging well-documented, organized workflows. We consider how established epidemiologists and early-career researchers alike can help midwife a culture of open science in epidemiology through their research practices, mentorship, and editorial activities.

Entwined life events: The effect of parental incarceration timing on children's academic achievement

Parental incarceration has negative effects on children's educational outcomes. Past studies have only analyzed, and therefore only treated as consequential, parental incarceration that occurs during childhood rather than prenatally. Such analyses that emphasize the importance only of events that occur during one's lifetime are common in life course studies. This paper introduces an "entwined life events" perspective, which argues that certain events are so consequential to multiple persons' lives that they should be analyzed as events within multiple independent life courses; parental incarceration, whenever it occurs, is entwined across and shapes both parents' and children's lives. Drawing on data from the Panel Study of Income Dynamics and the Fragile Families and Child Wellbeing Study, we find that parental incarceration, both prenatal and during childhood, significantly influences children's academic ability measures and years of completed schooling. Our results show heterogeneous effects by children's race. We find that the absolute magnitude of parental incarceration effect estimates is largest for White children relative to estimates for Black and Hispanic children. At the same time, outcome levels tend to be poorer for Black and Hispanic children with parental incarceration experience. We explain this racial heterogeneity as confounded by the many other social disadvantages that non-White children encounter, resulting in the individual effect of parental incarceration not being extremely disruptive to their academic growth.

Change in body weight and risk of hypertension after switching from efavirenz to dolutegravir in adults living with HIV: evidence from routine care in Johannesburg, South Africa

BACKGROUND

The integrase strand transfer inhibitor (INSTI) dolutegravir is recommended in World Health Organization guidelines, but is associated with weight gain. We evaluated weight change in patients switching from efavirenz to dolutegravir in first-line antiretroviral therapy (ART) in Johannesburg, South Africa.

METHODS

We conducted a prospective cohort study of adults (≥16 years) of black African ancestry with HIV who initiated ART between January 2010-December 2020. Patients were propensity score-matched 1:1 (unexposed i.e. remaining on efavirenz: exposed i.e. switched from efavirenz to dolutegravir) on sex, age, months on ART, first ART regimen, haemoglobin, body mass index (BMI), blood pressure, viral load and CD4 count. We used linear regression to assess the effect of switching from efavirenz to dolutegravir on weight change and hypertension 12 months after exposure.

FINDINGS

We matched 794 patients switching to dolutegravir to 794 remaining on efavirenz. Exposed patients had a higher mean change in weight (1.78 kg; 95% confidence interval (CI):1.04,2.52 kg) from start of follow-up to 12 months vs. unexposed. We also found a 14.2 percentage point increase (95% CI: 10.6,17.7) in the risk of hypertension in those exposed to dolutegravir vs those that remained on efavirenz.

INTERPRETATION

In a real-world population, patients gained more weight and were at higher risk of hypertension after switching from efavirenz to dolutegravir than those remaining on efavirenz. Longer follow-up is needed, however, to determine if INSTI-associated weight gain is associated with changes in non-communicable disease risk over the long-term, or whether weight gain is sustained, as seen in clinical trials.

FUNDING

This study has been made possible by the generous support of the American People and the President's Emergency Plan for AIDS Relief (PEPFAR) through the United States Agency for International Development (USAID), under the terms of cooperative agreement cooperative Agreement 72067419CA00004. In addition to the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) 1K01MH105320-01A1.

Long-term aircraft noise exposure and risk of hypertension in postmenopausal women

BACKGROUND

Studies of the association between aircraft noise and hypertension are complicated by inadequate control for potential confounders and a lack of longitudinal assessments, and existing evidence is inconclusive.

OBJECTIVES

We evaluated the association between long-term aircraft noise exposure and risk of hypertension among post-menopausal women in the Women's Health Initiative Clinical Trials, an ongoing prospective U.S.

COHORT

METHODS

Day-night average (DNL) and night equivalent sound levels (Lnight) were modeled for 90 U.S. airports from 1995 to 2010 in 5-year intervals using the Aviation Environmental Design Tool and linked to participant geocoded addresses from 1993 to 2010. Participants with modeled exposures ≥45 A-weighted decibels (dB [A]) were considered exposed, and those outside of 45 dB(A) who also did not live in close proximity to unmodeled airports were considered unexposed. Hypertension was defined as systolic/diastolic blood pressure ≥140/90 mmHg or inventoried/self-reported antihypertensive medication use. Using time-varying Cox proportional hazards models, we estimated hazard ratios (HRs) for incident hypertension when exposed to DNL or Lnight ≥45 versus <45 dB(A), controlling for sociodemographic, behavioral, and environmental/contextual factors.

RESULTS/DISCUSSION

There were 18,783 participants with non-missing DNL exposure and 14,443 with non-missing Lnight exposure at risk of hypertension. In adjusted models, DNL and Lnight ≥45 db(A) were associated with HRs of 1.00 (95% confidence interval [CI]: 0.93, 1.08) and 1.06 (95%CI: 0.91, 1.24), respectively. There was no evidence supporting a positive exposure-response relationship, and findings were robust in sensitivity analyses. Indications of elevated risk were seen among certain subgroups, such as those living in areas with lower population density (HRinteraction: 0.84; 95%CI: 0.72, 0.98) or nitrogen dioxide concentrations (HRinteraction: 0.82; 95%CI: 0.71, 0.95), which may indicate lower ambient/road traffic noise. Our findings do not suggest a relationship between aircraft noise and incident hypertension among older women in the U.S., though associations in lower ambient noise settings merit further investigation.

Assessment of facility-level antiretroviral treatment patient status utilizing a national-level laboratory cohort: Toward an understanding of system-level tracking and clinic switching in South Africa

Background

Most estimates of HIV retention are derived at the clinic level through antiretroviral (ART) patient management systems, which capture ART clinic visit data, yet these cannot account for silent transfers across HIV treatment sites. Patient laboratory monitoring visits may also be observed in routinely collected laboratory data, which include ART monitoring tests such as CD4 count and HIV viral load, key to our work here.

Methods

In this analysis, we utilized the NHLS National HIV Cohort (a system-wide viewpoint) to investigate the accuracy of facility-level estimates of retention in care for adult patients accessing care (defined using clinic visit data on patients under ART recorded in an electronic patient management system) at Themba Lethu Clinic (TLC). Furthermore, we describe patterns of facility switching among all patients and those patients classified as lost to follow-up (LTFU) at the facility level.

Results

Of the 43,538 unique patients in the TLC dataset, we included 20,093 of 25,514 possible patient records (78.8%) in our analysis that were linked with the NHLS National Cohort, and we restricted the analytic sample to patients initiating ART between 1 January 2007 and 31 December 2017. Most (60%) patients were female, and the median age (IQR) at ART initiation was 37 (31-45) years. We found the laboratory records augmented retention estimates by a median of 860 additional active records (about 8% of all median active records across all years) from the facility viewpoint; this augmentation was more noticeable from the system-wide viewpoint, which added evidence of activity of about one-third of total active records in 2017. In 2017, we found 7.0% misclassification at the facility-level viewpoint, a gap which is potentially solvable through data integration/triangulation. We observed 1,134/20,093 (5.6%) silent transfers; these were noticeably more female and younger than the entire dataset. We also report the most common locations for clinic switching at a provincial level.

Discussion

Integration of multiple data sources has the potential to reduce the misclassification of patients as being lost to care and help understand situations where clinic switching is common. This may help in prioritizing interventions that would assist patients moving between clinics and hopefully contribute to services that normalize formal transfers and fewer silent transfers.

Increased all-cause mortality following occupational injury: a comparison of two states

OBJECTIVES

To measure the impact of lost-time occupational injuries on all-cause mortality in Washington State and, using the same data elements and study design, to determine whether the estimated impact was similar to previous estimates for New Mexico.

METHODS

We linked injuries in the Washington workers' compensation system with Social Security Administration data on earnings and mortality. We estimated Cox survival models of mortality for women and men with lost-time compared with medical-only injuries, adjusting for age, pre-injury earnings and industry. We used quantitative bias analysis to account for confounding by pre-injury smoking and obesity.

RESULTS

The estimated mortality HR was 1.24 for women (95% CI 1.21 to 1.28) and 1.22 for men (95% CI 1.20 to 1.24). After adjusting for unmeasured pre-injury smoking and obesity, the estimated HR for women was 1.10, 95% simulation interval (SI) 1.00 to 1.21; for men, it was 1.15, 95% SI 1.04 to 1.27.

CONCLUSIONS

All-cause mortality for Washington workers with lost-time injuries was higher than for those with medical-only injuries. Estimated HRs for Washington were consistent with those previously estimated for New Mexico, a less populous state with lower median wages and a different workers' compensation insurance mechanism. This suggests that the relationship between workplace injury and long-term mortality may be generalisable to other US states. These findings support greater efforts to enhance safety and to investigate factors that improve postinjury employment opportunities and long-term health. This association should be examined in additional locations, with different study conditions, or using additional data on pre-injury risk factors.

Causal and Associational Language in Observational Health Research: A Systematic Evaluation

We estimated the degree to which language used in the high-profile medical/public health/epidemiology literature implied causality using language linking exposures to outcomes and action recommendations; examined disconnects between language and recommendations; identified the most common linking phrases; and estimated how strongly linking phrases imply causality. We searched for and screened 1,170 articles from 18 high-profile journals (65 per journal) published from 2010-2019. Based on written framing and systematic guidance, 3 reviewers rated the degree of causality implied in abstracts and full text for exposure/outcome linking language and action recommendations. Reviewers rated the causal implication of exposure/outcome linking language as none (no causal implication) in 13.8%, weak in 34.2%, moderate in 33.2%, and strong in 18.7% of abstracts. The implied causality of action recommendations was higher than the implied causality of linking sentences for 44.5% or commensurate for 40.3% of articles. The most common linking word in abstracts was "associate" (45.7%). Reviewers' ratings of linking word roots were highly heterogeneous; over half of reviewers rated "association" as having at least some causal implication. This research undercuts the assumption that avoiding "causal" words leads to clarity of interpretation in medical research.

A Framework for Descriptive Epidemiology

In this paper, we propose a framework for thinking through the design and conduct of descriptive epidemiologic studies. A well-defined descriptive question aims to quantify and characterize some feature of the health of a population and must clearly state: 1) the target population, characterized by person and place, and anchored in time; 2) the outcome, event, or health state or characteristic; and 3) the measure of occurrence that will be used to summarize the outcome (e.g., incidence, prevalence, average time to event, etc.). Additionally, 4) any auxiliary variables will be prespecified and their roles as stratification factors (to characterize the outcome distribution) or nuisance variables (to be standardized over) will be stated. We illustrate application of this framework to describe the prevalence of viral suppression on December 31, 2019, among people living with human immunodeficiency virus (HIV) who had been linked to HIV care in the United States. Application of this framework highlights biases that may arise from missing data, especially 1) differences between the target population and the analytical sample; 2) measurement error; 3) competing events, late entries, loss to follow-up, and inappropriate interpretation of the chosen measure of outcome occurrence; and 4) inappropriate adjustment.

Cohort profile: the South African National Health Laboratory Service (NHLS) National HIV Cohort

PURPOSE

South Africa's National Health Laboratory Service (NHLS) National HIV Cohort was established in 2015 to facilitate monitoring, evaluation and research on South Africa's National HIV Treatment Programme. In South Africa, 84.8% of people living with HIV know their HIV status; 70.7% who know their status are on ART; and 87.4% on ART are virologically suppressed.

PARTICIPANTS

The NHLS National HIV Cohort includes the laboratory data of nearly all patients receiving HIV care in the public sector since April 2004. Patients are included in the cohort if they have received a CD4 count or HIV RNA viral load (VL) test. Using an anonymised unique patient identifier that we have developed and validated to linked test results, we observe patients prospectively through their laboratory results as they receive HIV care and treatment. Patients in HIV care are seen for laboratory monitoring every 6-12 months. Data collected include age, sex, facility location and test results for CD4 counts, VLs and laboratory tests used to screen for potential treatment complications.

FINDINGS TO DATE

From April 2004 to April 2018, 63 million CD4 count and VL tests were conducted at 5483 facilities. 12.6 million unique patients had at least one CD4 count or VL, indicating they had accessed HIV care, and 7.1 million patients had a VL test indicating they had started antiretroviral therapy. The creation of NHLS National HIV Cohort has enabled longitudinal research on all lab-monitored patients in South Africa's national HIV programme, including analyses of (1) patient health at presentation; (2) care outcomes such as 'CD4 recovery', 'retention in care' and 'viral resuppression'; (3) patterns of transfer and re-entry into care; (4) facility-level variation in care outcomes; and (5) impacts of policies and guideline changes.

FUTURE PLANS

Continuous updating of the cohort, integration with available clinical data, and expansion to include tuberculosis and other lab-monitored comorbidities.

Adolescent retention in HIV care within differentiated service-delivery models in sub-Saharan Africa

Adolescents and young people living with HIV are at risk of disengaging from HIV care at all stages of the care cascade. Differentiated models of care offer simplified HIV-service delivery options in the hope of improving treatment outcomes, including retention on antiretroviral therapy. However, it remains unclear how successful and widespread these models are for adolescents in sub-Saharan Africa, where the burden of HIV is the greatest. Very few differentiated models of care specifically targeted to adolescents can be found and this priority group are currently ineligible from several models that exist. Where differentiated care has been made available to adolescents, data on the implementation and effectiveness of these interventions remain scarce. Despite this scarcity of evidence on the effectiveness of differentiated care among adolescent populations, several interventions, particularly community-based groups with peer navigators or supporters, might have potential to increase the reach, effectiveness, and adoption of differentiated care in adolescent HIV-care programmes.

Relationship Between Level of American Football Playing and Diagnosis of Chronic Traumatic Encephalopathy in a Selection Bias Analysis

Chronic traumatic encephalopathy (CTE) is a neurodegenerative disease associated with exposure to repetitive head impacts such as those from American football. Our understanding of this association is based on research in autopsied brains, since CTE can only be diagnosed postmortem. Such studies are susceptible to selection bias, which needs to be accounted for to ensure a generalizable estimate of the association between repetitive head impacts and CTE. We evaluated the relationship between level of American football playing and CTE diagnosis after adjusting for selection bias. The sample included 290 deceased male former American football players who donated their brains to the Veterans Affairs-Boston University-Concussion Legacy Foundation (VA-BU-CLF) Brain Bank between 2008 and 2019. After adjustment for selection bias, college-level and professional football players had 2.38 (95% simulation interval (SI): 1.16, 5.94) and 2.47 (95% SI: 1.46, 4.79) times the risk of being diagnosed with CTE as high-school-level players, respectively; these estimates are larger than estimates with no selection bias adjustment. Since CTE is currently diagnosed only postmortem, we additionally provide plausible scenarios for CTE risk ratios for each level of play during the former players' lifetime. This study provides further evidence to support a dose-response relationship between American football playing and CTE.

Misconceptions About the Direction of Bias From Nondifferential Misclassification

Measurement error is pervasive in epidemiologic research. Epidemiologists often assume that mismeasurement of study variables is nondifferential with respect to other analytical variables and then rely on the heuristic that "nondifferential misclassification will bias estimates towards the null." However, there are many exceptions to the heuristic for which bias towards the null cannot be assumed. In this paper, we compile and characterize 7 exceptions to this rule and encourage analysts to take a more critical and nuanced approach to evaluating and discussing bias from nondifferential mismeasurement.

Illustrating How to Simulate Data From Directed Acyclic Graphs to Understand Epidemiologic Concepts

Simulation methods are a powerful set of tools that can allow researchers to better characterize phenomena from the real world. As such, the ability to simulate data represents a critical set of skills that epidemiologists should use to better understand epidemiologic concepts and ensure that they have the tools to continue to self-teach even when their formal instruction ends. Simulation methods are not always taught in epidemiology methods courses, whereas causal directed acyclic graphs (DAGs) often are. Therefore, this paper details an approach to building simulations from DAGs and provides examples and code for learning to perform simulations. We recommend using very simple DAGs to learn the procedures and code necessary to set up a simulation that builds on key concepts frequently of interest to epidemiologists (e.g., mediation, confounding bias, M bias). We believe that following this approach will allow epidemiologists to gain confidence with a critical skill set that may in turn have a positive impact on how they conduct future epidemiologic studies.

One Pill, Once a Day: Simplified Treatment Regimens and Retention in HIV Care

Simplified drug regimens may improve retention in care for persons with chronic diseases. In April 2013, South Africa adopted a once-daily single-pill human immunodeficiency virus (HIV) treatment regimen as the standard of care, replacing a multiple-pill regimen. Because the regimens had similar biological efficacy, the shift to single-pill therapy offered a real-world test of the impact of simplified drug-delivery mechanisms on patient behavior. Using a quasi-experimental regression discontinuity design, we assessed retention in care among patients starting HIV treatment just before and just after the guideline change. The study included 4,484 patients starting treatment at a large public sector clinic in Johannesburg, South Africa. The share of patients prescribed a single-pill regimen increased by over 40 percentage points between March and April 2013. Initiating treatment after the policy change was associated with 11.7-percentage-points' higher retention at 12 months (95% confidence interval: -2.2, 29.4). Findings were robust to different measures of retention, different bandwidths, and different statistical models. Patients starting treatment early in HIV infection-a key population in the test-and-treat era-experienced the greatest improvements in retention from single-pill regimens.

The Expert Next Door: Interactions With Friends and Family During the COVID-19 Pandemic

The coronavirus disease 2019 (COVID-19) pandemic thrust the field of public health into the spotlight. For many epidemiologists, biostatisticians, and other public health professionals, this caused the professional aspects of our lives to collide with the personal, as friends and family reached out with concerns and questions. Learning how to navigate this space was new for many and required refining our communication depending on context, setting, and audience. Some of us took to social media, utilizing our existing personal accounts to share information after sorting through and summarizing the rapidly emerging literature to keep loved ones safe. However, those in our lives sometimes asked unanswerable questions, or began distancing themselves when we suggested more stringent guidance than they hoped, causing additional stress during an already traumatic time. We often had to remind ourselves that we are also individuals experiencing this pandemic, and that our time-intensive efforts were meaningful, relevant, and impactful. As this pandemic and other public health crises continue, we encourage our discipline to consider how we can best use shared lessons from this period, and recognize that our professional knowledge, when used in our personal lives, can promote, protect, and bolster confidence in public health.

Minimally Invasive Esophagectomy Is Associated with Superior Survival Compared to Open Surgery

INTRODUCTION

Minimally invasive esophagectomy (MIE) has not been associated with a long-term survival advantage compared to open esophagectomy (OE). We investigated survival differences between MIE, including laparoscopic and robotic, and OE.

METHODS

Patients undergoing esophagectomy from 2010 to 2014 with T_1-4N_0-3M₀, adenocarcinoma or squamous cell histology, in middle or lower esophagus were queried from the National Cancer Database and stratified into groups based on their surgical procedure: robotic, laparoscopic, or OE. Propensity matching (1:1) was done between robotic and laparoscopic to produce an MIE group. The MIE group was matched to OE yielding a 1:1:2 matching of robotic:laparoscopic:OE. Postoperative outcomes and survival (Kaplan-Meier) were compared between groups.

RESULTS

Prior to matching, 7,163 patients met inclusion criteria and a greater portion underwent OE (67.7%) than MIE (laparoscopic 24.9% and robotic 7.4%). Matching yielded similar groups (robotic = 527, laparoscopic = 527, and OE =1054). Compared to OE, MIE patients had a significantly greater number of nodes sampled and trended toward increased R0 resections (96.1% vs 94.3%, P = .053). OE was associated with a longer median postoperative stay (10 vs 9 days, P = .001). Mortality at 30 and 90 days was similar. However, postoperative survival for MIE was significantly greater than OE (P < .001). No survival difference existed between robotic and laparoscopic (P = .723).

CONCLUSIONS

MIE is associated with increased number of nodes examined and a shorter postoperative length of stay. After propensity matching, patients undergoing MIE had better long but not short-term survival than OE. This benefit seems to be independent of the use of robotic technology.

On the Need to Revitalize Descriptive Epidemiology

Nearly every introductory epidemiology course begins with a focus on person, place, and time, the key components of descriptive epidemiology. And yet in our experience, introductory epidemiology courses were the last time we spent any significant amount of training time focused on descriptive epidemiology. This gave us the impression that descriptive epidemiology does not suffer from bias and is less impactful than causal epidemiology. Descriptive epidemiology may also suffer from a lack of prestige in academia and may be more difficult to fund. We believe this does a disservice to the field and slows progress towards goals of improving population health and ensuring equity in health. The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) outbreak and subsequent coronavirus disease 2019 pandemic have highlighted the importance of descriptive epidemiology in responding to serious public health crises. In this commentary, we make the case for renewed focus on the importance of descriptive epidemiology in the epidemiology curriculum using SARS-CoV-2 as a motivating example. The framework for error we use in etiological research can be applied in descriptive research to focus on both systematic and random error. We use the current pandemic to illustrate differences between causal and descriptive epidemiology and areas where descriptive epidemiology can have an important impact.

The confounder matrix: A tool to assess confounding bias in systematic reviews of observational studies of etiology

Systematic reviews and meta-analyses are essential for drawing conclusions regarding etiologic associations between exposures or interventions and health outcomes. Observational studies comprise a substantive source of the evidence base. One major threat to their validity is residual confounding, which may occur when component studies adjust for different sets of confounders, fail to control for important confounders, or have classification errors resulting in only partial control of measured confounders. We present the confounder matrix-an approach for defining and summarizing adequate confounding control in systematic reviews of observational studies and incorporating this assessment into meta-analyses. First, an expert group reaches consensus regarding the core confounders that should be controlled and the best available method for their measurement. Second, a matrix graphically depicts how each component study accounted for each confounder. Third, the assessment of control adequacy informs quantitative synthesis. We illustrate the approach with studies of the association between short interpregnancy intervals and preterm birth. Our findings suggest that uncontrolled confounding, notably by reproductive history and sociodemographics, resulted in exaggerated estimates. Moreover, no studies adequately controlled for all core confounders, so we suspect residual confounding is present, even among studies with better control. The confounder matrix serves as an extension of previously published methodological guidance for observational research synthesis, enabling transparent reporting of confounding control and directly informing meta-analysis so that conclusions are drawn from the best available evidence. Widespread application could raise awareness about gaps across a body of work and allow for more valid inference with respect to confounder control.

Validation of Self-reported Opioid Agonist Treatment Among People Who Inject Drugs Using Prescription Dispensation Records

BACKGROUND

Studies of people who inject drugs (PWID) commonly use questionnaires to determine whether participants are currently, or have recently been, on opioid agonist treatment for opioid use disorder. However, these previously unvalidated self-reported treatment measures may be susceptible to inaccurate reporting.

METHODS

We linked baseline questionnaire data from 521 PWID in the Ontario integrated Supervised Injection Services cohort in Toronto (November 2018-March 2020) with record-level health administrative data. We assessed the validity (sensitivity, specificity, positive and negative predictive value [PPV and NPV]) of self-reported recent (in the past 6 months) and current (as of interview) opioid agonist treatment with methadone or buprenorphine-naloxone relative to prescription dispensation records from a provincial narcotics monitoring system, considered the reference standard.

RESULTS

For self-reported recent opioid agonist treatment, sensitivity was 78% (95% CI = 72, 83), specificity was 90% (95% CI = 86, 94), PPV was 90% (95% CI = 85, 93), and NPV was 79% (95% CI = 74, 84). For self-reported current opioid agonist treatment, sensitivity was 84% (95% CI = 78, 90), specificity was 87% (95% CI = 83, 91), PPV was 74% (95% CI = 67, 81), and NPV was 93% (95% CI = 89, 95).

CONCLUSIONS

Self-reported opioid agonist treatment measures were fairly accurate among PWID, with some exceptions. Inaccurate recall due to a lengthy lookback window may explain underreporting of recent treatment, whereas social desirability bias may have led to overreporting of current treatment. These validation data could be used in future studies of PWID to adjust for misclassification in similar self-reported treatment measures.

Potential for Selection Bias in Studies of the Association of Hormonal Contraception and Chronic Vulvar Pain

Background: Hormonal contraceptive use is common among reproductive-aged women, but research evaluating its etiological relationship to vulvodynia remains mixed. We sought to evaluate this association and examine the potential for bias due to care-seeking behavior. Materials and Methods: We conducted a case-control study of women recruited from a large health care network database from 2008 to 2011. Of 26,455 eligible respondents, 1168 met the case definition for chronic vulvar pain (CVP). We matched each case to three controls by age and used conditional logistic regression to calculate odds ratios (ORs) for prior hormonal contraception (HC) use and CVP, stratifying cases by whether or not they sought care for their vulvar pain. We also simulated the influence of potential biases due to care seeking, using parameters based on this dataset. Results: HC users had higher odds of CVP (adjusted OR = 2.6, 95% confidence interval [CI]: 2.2-3.2). Effect estimates were stronger when cases were restricted to care seekers (adjusted OR = 2.9, 95% CI: 2.2-3.7). Effect estimates decreased slightly as time increased between HC initiation and pain onset. Our simulations suggested that effect estimates may be spuriously strengthened when cases are restricted to care-seeking women, but controls are recruited from the general population. Conclusions: Our results suggest an association between antecedent HC use and CVP that is potentially spuriously strengthened in case-control studies when cases are restricted to care seekers but controls are not.

Virologic non-suppression and early loss to follow up among pregnant and non-pregnant adolescents aged 15-19 years initiating antiretroviral therapy in South Africa: a retrospective cohort study

Introduction

Older adolescents aged 15–19 years continue to have high rates of loss to follow up (LTFU), and high rates of virologic non‐suppression (VNS) compared to younger adolescents and adults. Adolescent females are at risk of pregnancy, which puts those living with HIV at a dual vulnerability. Our study assessed the factors associated with VNS and LTFU in older adolescents (including pregnant females) who initiated antiretroviral therapy (ART) in South Africa.

Methods

We included adolescents aged 15–19 years initiating ART between 2004 and 2019, with ≥ one viral load (VL) measurement between 4 and 24.5 months, and ≥ 6 months follow‐up, from six South African cohorts of the International epidemiology Databases to Evaluate AIDS‐Southern Africa (IeDEA‐SA). We defined VNS as VL ≥400 copies/ml and LTFU as not being in care for ≥180 days from ART start and not known as transferred out of the clinic or dead in the first 24 months on ART. We examined factors associated with VNS and LTFU using Fine&Gray competing risk models.

Results

We included a total of 2733 adolescents, 415 (15.2%) males, median (IQR) age at ART start of 18.6 (17.3, 19.4) years. Among females, 585/2318 (25.2%) were pregnant. Over the 24‐month follow‐up, 424 (15.5%) of all adolescents experienced VNS: range (11.1% pregnant females and 20.5% males). Over half of all adolescents were LTFU before any other event could occur. The hazard of VNS reduced with increasing age and CD4 count above 200 cells/μl at ART initiation among all adolescents having adjusted for all measured patient characteristics [adjusted sub‐distribution hazard ratio (aSHR) 19 vs. 15 years: 0.50 (95% CI: 0.36, 0.68), aSHR: >500 vs. ≤200 cells/μl: 0.22 (95% CI: 0.16, 0.31)]. The effect of CD4 count persisted in pregnant females. Increasing age and CD4 count >200 cells/μl were risk factors for LTFU among all adolescents.

Conclusions

Older adolescents had a high risk of LTFU shortly after ART start and a low risk of VNS, especially those initiating treatment during pregnancy. Interventions addressing adherence and retention should be incorporated into adolescent‐friendly services to prevent VNS and LTFU and endeavour to trace lost adolescents as soon as they are identified.