Unadjusted Estimates Research Articles

Inflation-adjusted analysis of ICER’s Unsupported Price Increase reports: contextualizing drug spending changes

Objective The Institute for Clinical and Economic Review (ICER) publishes annual Unsupported Price Increase (UPI) reports on prescription drugs, which have gained attention from policymakers and healthcare stakeholders. These reports do not adjust for inflation in their analyses of net price changes. This study aimed to evaluate the economic context of reported drug price increases in ICER’s UPI reports by applying inflation adjustments to the estimated impact on healthcare spending. Methods Separate inflation adjustments using the Medical Consumer Price Index (Medical CPI) and the Consumer Price Index for All Urban Consumers (CPI-U) were applied to unadjusted UPI findings. The study analyzed data from four most recent UPI reports spanning pricing data from 2018 to 2019 (report year 2020) to 2021 to 2022 (report year 2023), focusing on drugs classified as having “unsupported” price increases. Results Inflation-adjusted analyses showed lower estimates of real changes in spending compared to ICER’s unadjusted figures across all drugs analyzed. Across the four UPI reports, the average report-year difference in Medical CPI inflation adjustment spending versus unadjusted findings was 53%, whereas the average report-year difference in CPI-U inflation adjustment spending versus unadjusted findings was 80%. In the most recent report year of 2023, ICER’s unadjusted estimate of a $1.3 billion increase in healthcare spending transformed to a $134 million decrease when adjusted by Medical CPI and a $1.51 billion decrease when adjusted by CPI-U. Conclusion This study demonstrates that inflation adjustment can alter not only the findings but also the interpretation of drug spending changes in UPI reports. The findings highlight the importance of considering economic context in evaluating drug pricing trends and suggest that unadjusted figures may overstate the impact of real price changes. Given the influence of these reports on policy discussions, incorporating inflation adjustments into analyses is one of many steps toward informing evidence-based policy decisions.

Source-specific bias correction of US background and anthropogenic ozone modeled in CMAQ

Abstract. United States (US) background ozone (O3) is the counterfactual O3 that would exist with zero US anthropogenic emissions. Estimates of US background O3 typically come from chemical transport models (CTMs), but different models vary in their estimates of both background and total O3. Here, a measurement–model data fusion approach is used to estimate CTM biases in US anthropogenic O3 and multiple US background O3 sources, including natural emissions, long-range international emissions, short-range international emissions from Canada and Mexico, and stratospheric O3. Spatially and temporally varying bias correction factors adjust each simulated O3 component so that the sum of the adjusted components evaluates better against observations compared to unadjusted estimates. The estimated correction factors suggest a seasonally consistent positive bias in US anthropogenic O3 in the eastern US, with the bias becoming higher with coarser model resolution and with higher simulated total O3, though the bias does not increase much with higher observed O3. Summer average US anthropogenic O3 in the eastern US was estimated to be biased high by 2, 7, and 11 ppb (11 %, 32 %, and 49 %) for one set of simulations at 12, 36, and 108 km resolutions and 1 and 6 ppb (10 % and 37 %) for another set of simulations at 12 and 108 km resolutions. Correlation among different US background O3 components can increase the uncertainty in the estimation of the source-specific adjustment factors. Despite this, results indicate a negative bias in modeled estimates of the impact of stratospheric O3 at the surface, with a western US spring average bias of −3.5 ppb (−25 %) estimated based on a stratospheric O3 tracer. This type of data fusion approach can be extended to include data from multiple models to leverage the strengths of different data sources while reducing uncertainty in the US background ozone estimates.

Bias and negative values of COVID-19 vaccine effectiveness estimates from a test-negative design without controlling for prior SARS-CoV-2 infection

Test-negative designs (TNDs) are used to assess vaccine effectiveness (VE). Protection from infection-induced immunity may confound the association between case and vaccination status, but collecting reliable infection history can be challenging. If vaccinated individuals have less infection-induced protection than unvaccinated individuals, failure to account for infection history could underestimate VE, though the bias is not well understood. We simulated individual-level SARS-CoV-2 infection and COVID-19 vaccination histories and a TND. VE against symptomatic infection and VE against severe disease estimates unadjusted for infection history underestimated VE compared to estimates adjusted for infection history, and unadjusted estimates were more likely to be below 0%, which could lead to an incorrect interpretation that COVID-19 vaccines are harmful. TNDs assessing VE immediately following vaccine rollout introduced the largest bias and potential for negative VE against symptomatic infection. Despite the potential for bias, VE estimates from TNDs without prior infection information are useful because underestimation is rarely more than 8 percentage points.

Day Care Attendance and Risk of Type 1 Diabetes

A meta-analysis published in 2001 suggested that exposure to infections measured by day care attendance may be important in the pathogenesis of type 1 diabetes. Several new studies on the topic have since been published. To investigate the association between day care attendance and risk of type 1 diabetes and to include all available literature up to March 10, 2024. Data from PubMed and Web of Science were used and supplemented by bibliographies of the retrieved articles and searched for studies assessing the association between day care attendance and risk of type 1 diabetes. Studies that reported a measure of association between day care attendance and risk of type 1 diabetes were included. Details, including exposure and outcome assessment and adjustment for confounders, were extracted from the included studies. The multivariable association with the highest number of covariates, lowest number of covariates, and unadjusted estimates and corresponding 95% CIs were extracted. DerSimonian and Laird random-effects meta-analyses were performed and yielded conservative confidence intervals around relative risks. The principal association measure was day care attendance vs no day care attendance and risk of type 1 diabetes. Seventeen articles including 22 observational studies of 100 575 participants were included in the meta-analysis. Among the participants, 3693 had type 1 diabetes and 96 882 were controls. An inverse association between day care attendance and risk of type 1 diabetes was found (combined odds ratio, 0.68; 95% CI, 0.58-0.79; P < .001; adjusted for all available confounders). When the 3 cohort studies included were analyzed separately, the risk of type 1 diabetes was 15% lower in the group attending day care; however, the difference was not statistically significant (odds ratio, 0.85; 95% CI, 0.59-1.12; P = .37). These results demonstrated that day care attendance appears to be associated with a reduced risk of type 1 diabetes. Increased contacts with microbes in children attending day care compared with children who do not attend day care may explain these findings. However, further prospective cohort studies are needed to confirm the proposed association.

Effect of GLP-1 receptor agonists on prostate cancer risk reduction: a systematic review and meta-analysis.

Prostate cancer is one of the most prevalent malignancies among men globally. Glucagon-like peptide 1 receptor agonists (GLP-1 RAs), primarily used for type 2 diabetes mellitus (T2DM) management, have been investigated for their potential effects on cancer risks. This systematic review and meta-analysis aimed to assess the association between GLP-1 RA use and risk reduction of prostate cancer. A comprehensive literature search was conducted across PubMed, Embase, and Web of Science up to July 30, 2024. Studies that met the inclusion criteria randomized controlled trials, cohort studies, case-control studies, and observational studies assessing the incidence of prostate cancer in GLP-1 RA-treated patients were included. The quality of studies was evaluated using the Newcastle-Ottawa Scale and the Cochrane Risk of Bias tool. Meta-analysis was performed using a random effects model. A total of five studies were included, analyzing data from diverse international contexts. The included studies showed a reduced risk of prostate cancer with both adjusted and unadjusted effect estimates with GLP-1 RAs. The meta-analysis revealed an RR of 0.72 (95% CI: 0.610 to 0.832), indicating a statistically significant 28% reduction in prostate cancer risk associated with GLP-1 RA use compared to placebo or other antidiabetic drugs. Moderate heterogeneity was observed (I2 = 51%). Sensitivity analysis confirmed the results. The findings suggest a significant protective association between GLP-1 RA use and reduced prostate cancer risk in men, particularly those with T2DM. This supports the potential of GLP-1 RAs not only in diabetes management but also as a strategy to mitigate cancer risk. Further research is required to confirm these findings and explore the underlying mechanisms, considering different dosages, durations of therapy, and patient subgroups based on demographic and metabolic characteristics.

Intraoperative neurophysiologic monitoring during cardiac surgery: an observational cohort study

ObjectiveTo evaluate the impact of intraoperative neuromonitoring (IONM) on stroke and operative mortality after coronary and/or valvular operations.MethodsThis was an observational study of coronary and/or valvular heart operations from 2010...

When does adjusting covariate under randomization help? A comparative study on current practices

PurposeWe aim to thoroughly compare past and current methods that leverage baseline covariate information to estimate the average treatment effect (ATE) using data from of randomized clinical trials (RCTs). We especially focus on their performance, efficiency gain, and power.MethodsWe compared 6 different methods using extensive Monte-Carlo simulation studies: the unadjusted estimator, i.e., analysis of variance (ANOVA), the analysis of covariance (ANCOVA), the analysis of heterogeneous covariance (ANHECOVA), the inverse probability weighting (IPW), the augmented inverse probability weighting (AIPW), and the overlap weighting (OW) as well as the augmented overlap weighting (AOW) estimators. The performance of these methods is assessed using the relative bias (RB), the root mean square error (RMSE), the model-based standard error (SE) estimation, the coverage probability (CP), and the statistical power.ResultsEven with a well-executed randomization, adjusting for baseline covariates by an appropriate method can be a good practice. When the outcome model(s) used in a covariate-adjusted method is closer to the correctly specified model(s), the efficiency and power gained can be substantial. We also found that most covariate-adjusted methods can suffer from the high-dimensional curse, i.e., when the number of covariates is relatively high compared to the sample size, they can have poor performance (along with lower efficiency) in estimating ATE. Among the different methods we compared, the OW performs the best overall with smaller RMSEs and smaller model-based SEs, which also result in higher power when the true effect is non-zero. Furthermore, the OW is more robust when dealing with the high-dimensional issue.ConclusionTo effectively use covariate adjustment methods, understanding their nature is important for practical investigators. Our study shows that outcome model misspecification and high-dimension are two main burdens in a covariate adjustment method to gain higher efficiency and power. When these factors are appropriately considered, e.g., performing some variable selections if the data dimension is high before adjusting covariate, these methods are expected to be useful.

Type and dimensions can predict ventricular arrhythmias and cardiac death in primary benign cardiac tumors in children

Despite being extremely rare in children, primary benign cardiac tumors can cause malignant ventricular arrhythmias (VA) or even sudden cardiac death. To assess the predictors of cardiovascular death and malignant VAs, we designed a retrospective single-center study enrolling paediatric patients. We defined as primary outcome a composite of cardiovascular death, sustained VT, ventricular fibrillation and rapid, symptomatic non-sustained VT. Our secondary endpoint was to assess the prevalence of clinically significant arrhythmias in our population. Methods and resultsWe fitted a multivariate Cox regression model to assess the predictors of the primary outcome. Over a period of 38 years, a total of 97 children were enrolled in the study. Among them, there were 73 rhabdomyomas, 13 fibromas, 3 myxomas, 3 teratomas, 1 lipoma, 2 haemangiomas and 2 fibroelastomas. Over a median follow up of 10.53 years, 16 patients met the primary outcome. Kaplan Meier unadjusted survival estimates showed that tumor dimensions larger than 2.3 cm and diagnosis of fibroma predicted worse outcomes compared with smaller tumors or other histotypes, (log rank p < 0.0002 and < 0.0001 respectively). In multivariate Cox proportional hazards analysis, diagnosis of fibroma and tumor dimensions were independently associated to the primary endpoint (HR: 5.06, 95 %CI (1.3–19); and 1.26 • (1.05–11), respectively). Clinically significant arrhythmias were reported in 24.5 % of the study population. ConclusionsAmong paediatric primary cardiac tumors, type and dimensions may predict the hazard of malignant VAs and cardiac death.

Special educational needs provision in primary education and its impact on unauthorised absence in children born with Cleft Lip and Palate: a population level study using linked health and education data

IntroductionWithout accounting for the underlying need for Special Educational Needs (SEN) provision, pupils with SEN in England have been shown to have higher rates of unauthorised absences. Objectives and ApproachOur objective was to understand the impact of SEN provision on unauthorised absences in children with an expected need for SEN, those born with Cleft Lip and/or Palate. We used linked health and educational records to investigate the relationship between three categories of SEN provision (No Provision versus SEN Support versus Education and Healthcare Plan (EHCP)) and unauthorised absences. We aimed to causally estimate the incident rate ratio (IRR) of SEN Support vs No Provision and EHCP vs SEN support on unauthorised absences between years one and six of primary education. Combining sociodemographic, educational and clinical variables from linked data, we used propensity-score based methods to adjust for non-randomised assignment to these categories of treatment. ConclusionsBetween academic years 2009 to 2019, we found common support between comparison groups No provision (n = 4350) and SEN support (n = 2009) only for children with Cleft lip and palate and only reported on these comparisons. Unadjusted estimates demonstrate a 22% increase in absences in children with SEN support over No Support (IRR: 1.22 95%CI: 1.10-1.34). Reweighting for need using data from linked health and education records demonstrated a 14% reduction in unauthorised absences (IRR: 0.86 95%CI: 0.76-0.97). ImplicationContrary to prior findings, by carefully matching on detailed linked health and education characteristics, we find SEN provision reduces unauthorised absences

Tyranny-of-the-Minority Regression Adjustment in Randomized Experiments

Regression adjustment is widely used in the analysis of randomized experiments to improve the estimation efficiency of the treatment effect. This article reexamines a weighted regression adjustment method termed tyranny-of-the-minority (ToM), wherein units in the minority group are given greater weights. We demonstrate that ToM regression adjustment is more robust than Lin’s regression adjustment with treatment-covariate interactions, even though these two regression adjustment methods are asymptotically equivalent in completely randomized experiments. Moreover, ToM regression adjustment can be easily extended to stratified randomized experiments and completely randomized survey experiments. We obtain the design-based properties of the ToM regression-adjusted average treatment effect estimator under such designs. In particular, we show that the ToM regression-adjusted estimator improves the asymptotic estimation efficiency compared to the unadjusted estimator, even when the regression model is misspecified, and is optimal in the class of linearly adjusted estimators. We also study the asymptotic properties of various heteroscedasticity-robust standard errors and provide recommendations for practitioners. Simulation studies and real data analysis demonstrate ToM regression adjustment’s superiority over existing methods. Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.

Inequities in medicaid home- and community-based services waiver enrollment among people with intellectual and/or developmental disabilities: A nationwide claims-based analysis

BackgroundStates use Medicaid 1915(c) waiver programs to enable access to home- and community-based services for people with intellectual and/or developmental disabilities (I/DD). However, enrollment rates and potential inequities are not well documented, impeding efforts to improve care access and quality for waiver program enrollees, especially for racially minoritized beneficiaries experiencing compounded barriers to services and supports. ObjectiveTo characterize year-by-year 1915(c) waiver program enrollment among Medicaid-enrolled adults with I/DD from 2016 to 2019 and to analyze population-level inequities by type of I/DD and racial/ethnic group. MethodsOur data source was 2016–2019 Medicaid Transformed Medicaid Statistical Information System Analytic Files Demographic and Eligibility files for beneficiaries with Down syndrome, autism, and intellectual disability. We used generalized estimating equation linear models to estimate the associations of type of I/DD and racial/ethnic group with the probability of 1915(c) waiver program enrollment and reported (1) unadjusted estimates and (2) estimates adjusted for demographics with state and year fixed effects. ResultsFrom 2016 to 2019, across all types of I/DD and racial/ethnic groups, unadjusted 1915(c) waiver program enrollment rates ranged from 40 to 60 % nationwide. We found modest growth in 1915(c) I/DD waiver program enrollment but persistent inequities over time. Compared to beneficiaries with intellectual disabilities, beneficiaries with autism were less likely to enroll while beneficiaries with Down syndrome were more likely. While some racial/ethnic groups had higher unadjusted mean enrollment, after adjustment, racially minoritized beneficiaries were 3.66–12.0 percentage points less likely to enroll compared to white non-Hispanic beneficiaries. ConclusionsGiven extensive waiting lists for 1915(c) waiver programs, Medicaid programs should evaluate existing enrollment and authorization processes and consider alternative HCBS program authorities.

Real-world outcomes of lobectomy, segmentectomy and wedge resection for the treatment of stage c-IA lung carcinoma.

To determine safety and survival outcomes associated with lobectomy, segmentectomy and wedge resection for early-stage lung cancer by quiring the French population-based registry EPIdemiology in THORacic surgery (EPITHOR). Retrospective analysis of 19 452 patients with stage c IA lung carcinoma who underwent lobectomy, segmentectomy or wedge resection between 2016 and 2022 with curative-intent. Main outcome measures were 90-day mortality and 5-year overall survival estimates. Proportional hazards regression and propensity score matching were used to adjust outcomes for key patient, tumour and practice environment factors. The treatment distribution was 72.2% for lobectomy, 21.5% for segmentectomy and 6.3% for wedge. Unadjusted 90-day mortality rates were 1.6%, 1.2% and 1.1%, respectively (P = 0.10). Unadjusted 5-year overall survival estimates were 80%, 78% and 70%, with significant inter-group survival curves differences (P < 0.0001). Multivariable proportional hazards regression showed that wedge was associated with worse overall survival [adjusted hazard ratio (AHR), 1.23 (95% confidence interval 1.03-1.47); P = 0.021] compared with lobectomy, while no significant difference was disclosed when comparing segmentectomy to lobectomy (1.08 [0.97-1.20]; P = 0.162). The three-way propensity score analyses confirmed similar 90-day mortality rate for wedge resection and segmentectomy compared with lobectomy (hazard ratio: 0.43; 95% confidence interval 0.16-1.11; P = 0.081 and 0.99; 0.48-2.10; P = 0.998, respectively), but poorer overall survival (1.45; 1.13-1.86; P = 0.003 and 1.31; 1-1.71; P = 0.048, respectively). Wedge resection was associated with comparable 90-day mortality but lower overall survival when compared to lobectomy. Overall, all types of sublobar resections may not offer equivalent oncologic effectiveness in real-world settings.

An Updated Data Portrait of Heterosexual, Gay/Lesbian, Bisexual, and Other Sexual Minorities in the United States

Sexual minorities are a rapidly growing population, with recent estimates showing a two-fold increase in the percentage of sexual minorities over the past decade. Working with relatively few measures to identify sexual minorities, social scientists have amassed an impressive amount of evidence on inequality by sexuality. Despite this remarkable work, I argue that it is important to take a step back analytically and re-assess sexual minorities from a descriptive standpoint. Using population-level data from the CDC’s Behavioral Risk Factor Surveillance System, I provide unadjusted estimates of sociodemographic, socioeconomic, and family characteristics by sexual identity. Results reveal that sexual minorities are younger, are more racially diverse, and concentrate in different parts of the country than heterosexuals. Similarly, sexual minorities have remarkably different socioeconomic lives than heterosexuals, who enjoy higher annual household incomes, achieve higher educational attainment, and are more likely to be homeowners. Sexual minorities are also less likely to be married than heterosexuals. I conclude by highlighting that descriptive research can illuminate compositional differences between sexual minorities and heterosexuals; provide rationales for adjusting for certain characteristics that might confound relationships between sexual identity and numerous outcomes; and highlight potential explanatory mechanisms to make better sense of well-established findings regarding sexual minority disadvantage.

Prognostic Factors for Early Recurrence After Resection of Pancreatic Cancer: A Systematic Review and Meta-Analysis

More than half of pancreatic ductal adenocarcinomas (PDACs) recur within 12 months after curative-intent resection. This systematic review and meta-analysis was conducted to identify all reported prognostic factors for early recurrence in resected PDACs. After a systematic literature search, a meta-analysis was conducted using a random effects model. Separate analyses were performed for adjusted vs unadjusted effect estimates as well as reported odds ratios (ORs) and hazard ratios (HRs). Risk of bias was assessed using the Quality in Prognostic Studies tool, and evidence was rated according to Grading of Recommendations Assessment, Development and Evaluation recommendations. After 2903 abstracts were screened, 65 studies were included. Of these, 28 studies (43.1%) defined early recurrence as evidence of recurrence within 6 months, whereas 34 (52.3%) defined it as evidence of recurrence within 12 months after surgery. Other definitions were uncommon. Analysis of unadjusted ORs and HRs revealed 41 and 5 prognostic factors for early recurrence within 6 months, respectively. When exclusively considering adjusted data, we identified 25 and 10 prognostic factors based on OR and HR, respectively. Using a 12-month definition, we identified 38 (OR) and 15 (HR) prognostic factors from unadjusted data and 38 (OR) and 30 (HR) prognostic factors from adjusted data, respectively. On the basis of frequency counts of adjusted data, preoperative carbohydrate antigen 19-9, N status, nondelivery of adjuvant therapy, grading, and tumor size based on imaging were identified as key prognostic factors for early recurrence. Reported prognostic factors of early recurrence vary considerably. Identified key prognostic factors could aid in the development of a risk stratification framework for early recurrence. However, prospective validation is necessary.

Association between breastfeeding duration and educational attainment in rural Southwest Uganda: a population-based cohort study

ABSTRACT Background Breastfeeding is important for early childhood nutrition and health. The positive effects on educational outcomes may be attributed to socioeconomic factors. Socioeconomic status is not a strong predictor of breastfeeding in sub-Saharan African countries. Yet, few studies have investigated the association between breastfeeding and educational outcomes in these countries. Objective This study investigated the association between breastfeeding duration and children’s educational attainment in rural Southwest Uganda. Methods We analysed longitudinal data on 3018 children who had information on breastfeeding and were followed for at least 5 years, with at least one primary school grade recorded by 2005. Data on breastfeeding duration were collected from mothers. The highest school grade was recorded repeatedly between ages 6 and 12 years. We calculated age-for-grade based on whether a child was on, over, or under the official age for a grade. Generalised estimating equations and binary logistic regression estimated the effect of breastfeeding duration on being 2 years, 3 or more years, or any years over-age for grade in primary school, adjusting for socioeconomic status and maternal-child characteristics. Results Most mothers breastfed for more than a year. Just over one-third breastfed for 18–23 months, and 30% breastfed for longer. By age eight, 42% of the children were two years over-age for their grade. Three or more years over-age for grade increased from 19% at age nine to 56% at age 12. Both adjusted and unadjusted estimates were consistent in showing reduced odds for children being 2 years, 3 or more years, or any years over-age for grade among children breastfed for 7–12, 13–17, 18–23, and > 23 months compared to those breastfed for 0–6 months. There was no evidence to support an overall association between breastfeeding duration and being over-age for grade. There was no evidence of association in the sex and age sub-group analyses. Conclusion Although we found no association between breastfeeding duration and educational attainment, breastfeeding remains important for children’s health and nutrition, and mothers should be encouraged and supported to breastfeed for the recommended duration.

Transcatheter mitral valve repair with MitraClip: comparison of NT, NTr, and XTr Devices.

Transcatheter edge-to-edge repair (TEER) has become an established minimally invasive treatment for significant mitral regurgitation. Ongoing refinements and the availability of different clipping devices have expanded the indications for and effectiveness of TEER, but comprehensive comparative data on this issue are lacking. In this study, we compared NT, NTr, and XTr MitraClip devices (Abbot) for TEER. Details on patient, imaging, and procedural details, as well as short- and long-term outcomes, were sought from a national prospective clinical registry on TEER with MitraClip. The primary outcome of interest was discharge after procedural success without major clinical complications. A total of 2236 patients were included, 1228 (54.9%) in whom NT implantation only was attempted, 233 (10.4%) in whom NTr but not XTr implantation was attempted, and 775 (34.7%) in whom XTr implantation was attempted. Clinical and imaging features differed substantially across the groups, reflecting expanding indications with NTr and XTr devices. In-hospital outcomes were largely similar among the 3 groups, including death. Long-term unadjusted estimates of effect showed significant differences in several outcomes, including death, rehospitalization, and their composite, which demonstrated that NT was associated with more unfavorable outcomes compared with the other devices (all P less than .05). However, most differences depended on baseline features, as adjusted analysis showed no significant differences for early as well as long-term outcomes, including long-term death, rehospitalization, and their composite (all P greater than .05). New-generation MitraClip devices are associated with favorable procedural and clinical outcomes, despite being used in patients with more adverse features, when compared with patients treated with previous devices.

Prevalence of long COVID symptoms in Haryana, India: a cross-sectional follow-up study

Emerging research indicates growing concern over long COVID globally, although there have been limited studies that estimate population burden. We aimed to estimate the burden of long COVID in three districts of Haryana, India, using an opportunity to link a seroprevalence study to follow-up survey of symptoms associated with long COVID. We used a population-based seroprevalence survey for COVID-19 conducted in September 2021 across Haryana, India. Adults from three purposively selected districts (Rohtak, Gurugram, and Jhajjar) were eligible to participate; 2205 of 3213 consented to participate in a survey on health status. Trained investigators administered a structured questionnaire that included demographic characteristics, self-reported symptoms of illness in the last six months before the survey, mental health, and history of COVID-19. Unadjusted regression estimates indicated positive correlations between symptomatic complaints and COVID-19 exposure, suggesting lingering effects of COVID-19 in this population. The overall physical morbidity index was higher among those who tested positive for COVID-19, as was the incidence of new cases. However, both morbidity and incidence became statistically insignificant after adjustment for multiple comparisons. Cough emerged as the only statistically significant individual persistent symptom. Sex-stratified analyses indicated significant estimates only for physical morbidity in women. This study is one of the first from India that uses a large population-based sample to examine longer term repercussions of COVID infections. The burden of long COVID should primarily be addressed in clinical settings, where specialised treatment for individual cases continues to evolve. Our analyses also provide insight into the size and nature of studies required to assess the population-level burden of long COVID. This paper was produced under the auspices of the Lancet COVID 19 Commission India Task Force, which was supported financially by the Reliance Foundation. The Lancet COVID 19 Commission was set up in July 2020 and submitted its final report by October 2022. This report by the India Task Force was prepared during the same period.

Centre-level variation in the survival of patients receiving haemodialysis in India: findings from a nationwide private haemodialysis network

There are no large studies examining survival in patients receiving haemodialysis in India or considering centre-level effects on survival. We measured survival variation between dialysis centres across India and evaluated the extent to which differences are explained by measured centre characteristics. This is a multilevel analysis of patient survival in centres of the NephroPlus dialysis network consisting of 193 centres across India. Patients receiving haemodialysis at a centre for ≥90 days between April 2014 and June 2019 were included, with analyses restricted to centres with ≥10 such patients. The primary outcome was all-cause mortality, measured from 90 days after joining a centre. Proportional hazards models with shared frailty were used to model centre- and patient-level effects on survival. Amongst 23,601 patients (median age 53 years; 29% female), the unadjusted centre-specific 180-day Kaplan-Meier survival estimates ranged between 55% (95% confidence interval [CI] 38-80%) and 100%, with a median of 88% (interquartile interval 83%-92%). After accounting for multilevel factors, estimated 180-day survival ranged between 83% (73-89%) and 97% (95-98%), with 90% 180-day survival in the average centre. The mortality rate in patients attending rural centres was 32% (Hazard Ratio 1.32; 95% CI 1.06-1.65) higher than those at urban centres in adjusted analyses. Multiple patient characteristics were associated with mortality. This is the first national benchmark for survival amongst dialysis patients in India. Centre- and patient-level characteristics are associated with survival but there remains unexplained variation between centres. As India continues to widen dialysis access, ongoing quality improvement programs will be an important part of ensuring that patients experience the best possible outcomes at the point of care. This project received no external funding.

Mortality Among HIV-Infected Adults on Antiretroviral Therapy in Southern Uganda.

Monitoring and evaluation of clinical programs requires assessing patient outcomes. Numerous challenges complicate these efforts, the most insidious of which is loss to follow-up (LTFU). LTFU is a composite outcome, including individuals out of care, undocumented transfers, and unreported deaths. Incorporation of vital status information from routine patient outreach may improve the mortality estimates for those LTFU. We analyzed routinely collected clinical and patient tracing data for individuals (15 years or older) initiating antiretroviral treatment between January 2014 and December 2018 at 2 public HIV care clinics in greater Rakai, Uganda. We derived unadjusted mortality estimates using Kaplan-Meier methods. Estimates, adjusted for unreported deaths, applied weighting through the Frangakis and Rubin method to represent outcomes among LTFU patients who were successfully traced and for whom vital status was ascertained. Confidence intervals were determined through bootstrap methods. Of 1969 patients with median age at antiretroviral treatment initiation of 31 years (interquartile range: 25-38), 1126 (57.2%) were female patients and 808 (41%) were lost. Of the lost patients, 640 patient files (79.2%) were found and reviewed, of which 204 (31.8%) had a tracing attempt. Within the electronic health records of the program, 28 deaths were identified with an estimated unadjusted mortality 1 year after antiretroviral treatment initiation of 2.5% (95% CI: 1.8% to 3.3%). Using chart review and patient tracing data, an additional 24 deaths (total 52) were discovered with an adjusted 1-year mortality of 3.8% (95% CI: 2.6% to 5.0%). Data from routine outreach efforts by HIV care and treatment programs can be used to support plausible adjustments to estimates of client mortality. Mortality estimates without active ascertainment of vital status of LTFU patients may significantly underestimate program mortality.

Descriptive inference using large, unrepresentative nonprobability samples: An introduction for ecologists.

Biodiversity monitoring usually involves drawing inferences about some variable of interest across a defined landscape from observations made at a sample of locations within that landscape. If the variable of interest differs between sampled and nonsampled locations, and no mitigating action is taken, then the sample is unrepresentative and inferences drawn from it will be biased. It is possible to adjust unrepresentative samples so that they more closely resemble the wider landscape in terms of "auxiliary variables." A good auxiliary variable is a common cause of sample inclusion and the variable of interest, and if it explains an appreciable portion of the variance in both, then inferences drawn from the adjusted sample will be closer to the truth. We applied six types of survey sample adjustment-subsampling, quasirandomization, poststratification, superpopulation modeling, a "doubly robust" procedure, and multilevel regression and poststratification-to a simple two-part biodiversity monitoring problem. The first part was to estimate the mean occupancy of the plant Calluna vulgaris in Great Britain in two time periods (1987-1999 and 2010-2019); the second was to estimate the difference between the two (i.e., the trend). We estimated the means and trend using large, but (originally) unrepresentative, samples from a citizen science dataset. Compared with the unadjusted estimates, the means and trends estimated using most adjustment methods were more accurate, although standard uncertainty intervals generally did not cover the true values. Completely unbiased inference is not possible from an unrepresentative sample without knowing and having data on all relevant auxiliary variables. Adjustments can reduce the bias if auxiliary variables are available and selected carefully, but the potential for residual bias should be acknowledged and reported.