A Cautionary Note on Using Propensity Score Calibration to Control for Unmeasured Confounding Bias When the Surrogacy Assumption Is Absent.

Conventional propensity score methods encounter challenges when unmeasured confounding is present, as it becomes impossible to accurately estimate the gold-standard propensity score when data on certain confounders are unavailable. Propensity score calibration (PSC) addresses this issue by constructing a surrogate for the gold-standard propensity score under the surrogacy assumption. This assumption posits that the error-prone propensity score, based on observed confounders, is independent of the outcome when conditioned on the gold-standard propensity score and the exposure. However, this assumption implies that confounders cannot directly impact the outcome and that their effects on the outcome are solely mediated through the propensity score. This raises concerns regarding the applicability of PSC in practical settings where confounders can directly affect the outcome. While PSC aims to target a conditional treatment effect by conditioning on a subject's unobservable propensity score, the causal interest in the latter case lies in a conditional treatment effect conditioned on a subject's baseline characteristics. Our analysis reveals that PSC is generally biased unless the effects of confounders on the outcome and treatment are proportional to each other. Furthermore, we identify 2 sources of bias: 1) the noncollapsibility of effect measures, such as the odds ratio or hazard ratio and 2) residual confounding, as the calibrated propensity score may not possess the properties of a valid propensity score.

The Relationship Between Hospital Capability and Mortality in Sepsis: Development of a Sepsis-Related Hospital Capability Index.

OBJECTIVES: Regionalized sepsis care could improve sepsis outcomes by facilitating the interhospital transfer of patients to higher-capability hospitals. There are no measures of sepsis capability to guide the identification of such hospitals, although hospital case volume of sepsis has been used as a proxy. We evaluated the performance of a novel hospital sepsis-related capability (SRC) index as compared with sepsis case volume. DESIGN: Principal component analysis (PCA) and retrospective cohort study. SETTING: A total of 182 New York (derivation) and 274 Florida and Massachusetts (validation) nonfederal hospitals, 2018. PATIENTS: A total of 89,069 and 139,977 adult patients (≥ 18 yr) with sepsis were directly admitted into the derivation and validation cohort hospitals, respectively. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We derived SRC scores by PCA of six hospital resource use characteristics (bed capacity, annual volumes of sepsis, major diagnostic procedures, renal replacement therapy, mechanical ventilation, and major therapeutic procedures) and classified hospitals into capability score tertiles: high, intermediate, and low. High-capability hospitals were mostly urban teaching hospitals. Compared with sepsis volume, the SRC score explained more variation in hospital-level sepsis mortality in the derivation (unadjusted coefficient of determination [ R2 ]: 0.25 vs 0.12, p < 0.001 for both) and validation (0.18 vs 0.05, p < 0.001 for both) cohorts; and demonstrated stronger correlation with outward transfer rates for sepsis in the derivation (Spearman coefficient [ r ]: 0.60 vs 0.50) and validation (0.51 vs 0.45) cohorts. Compared with low-capability hospitals, patients with sepsis directly admitted into high-capability hospitals had a greater number of acute organ dysfunctions, a higher proportion of surgical hospitalizations, and higher adjusted mortality (odds ratio [OR], 1.55; 95% CI, 1.25-1.92). In stratified analysis, worse mortality associated with higher hospital capability was only evident among patients with three or more organ dysfunctions (OR, 1.88 [1.50-2.34]). CONCLUSIONS: The SRC score has face validity for capability-based groupings of hospitals. Sepsis care may already be de facto regionalized at high-capability hospitals. Low-capability hospitals may have become more adept at treating less complicated sepsis.

The impact of virtual interviews on the Fellowship Council application and main match process.

BACKGROUND: The COVID-19 pandemic led the Fellowship Council (FC) to transition rapidly from in-person to virtual interviews. We investigated the impact of this transition on the FC application and main match process. METHODS: Five years (2018-2022) of deidentified FC applicant, program, and match rank data were used to assess differences between in-person (2018-2019) and virtual interview (2021-2022) cycles. Data are expressed as mean ± SD and one-way and two-way MANOVA tests were applied. RESULTS: Trainees applied to an average of 30.4 ± 24.3 programs and ranked an average of 10.7 ± 9.7 programs with a 57% match rate and average rank position of 3.6 ± 3.3. Fellowship programs received an average of 64.9 ± 28.6 applications and ranked an average of 15.4 ± 8.8 applicants with a 95% match rate and average applicant rank position of 3.0 ± 3.4. Applicants who interviewed virtually applied to a greater number of programs (32.7 vs. 27.0; p < 0.001) and ranked a greater number of programs (11.5 vs. 10.0; p = 0.004) with no difference in match rates (58% vs. 55%, p = 0.291). Among matched applicants, there was a significant difference in average rank position (3.20 vs. 4.30, p < 0.001), favoring the in-person cohort. Fellowship programs had more applicants per program (69.2 vs. 57.8; p < 0.001) and ranked more applicants (17.4 vs. 13.3; p < 0.001) during the virtual interview cycles. No difference in either match rates (93% vs. 96%, p = 0.178) or applicant rank position (3.09 vs. 2.93, p = 0.561) was seen between in-person and virtual application cycles. CONCLUSION: Virtual interviews were associated with an increased number of applications for fellowship and applicants ranked by programs but did not impact match rates of either group. Rank match position declined somewhat for applicants but not for fellowship programs. Virtual interviews offer more opportunities for applicants and a greater number of candidates for fellowship programs with only a slight decrement in fellow match rank position.

Factors influencing the prevalence of cervical cancer screening in Malaysia: a nationwide survey.

BACKGROUND: In 2020, cervical cancer ranked fourth in terms of both frequency of diagnosis and the leading cause of cancer-related deaths among women globally. Among Malaysian women, it was the third most prevalent form of cancer. Published data on nationally representative cervical cancer screening in Malaysia have been limited. Therefore, this study aimed to determine the prevalence of receiving a Pap smear test in the past three years, its relationship with socio-demographic factors and physical activity. METHODS: Using a subset of survey data from the National Health and Morbidity Survey (NHMS) 2019, a secondary data analysis was performed. Trained research assistants collected data through face-to-face method using a mobile tablet questionnaire system application. Logistic regression analysis was performed to examine the relationship between sociodemographic factors, physical activity, and cervical cancer screening. The analyses were conducted using STATA version 14 (Stata Corp, College Station, Texas, USA), accounting for sample weighs and complex sampling design. RESULTS: The analysis included 5,650 female respondents, representing an estimated 10.3 million Malaysian female adults aged 18 and above. Overall, 35.2% (95%CI 33.2, 37.4) respondents had a Pap smear test within the past three years. Respondents who were physically active were 1.41 times more likely to have a Pap smear test. Similarly, respondents aged 35-59 (OR 1.84; 95%CI 1.46, 2.34) and those living in rural localities (OR 1.38; 95%CI 1.13, 1.70) had higher odds of receiving a Pap smear test. Compared to married respondents, single respondents (OR 0.04; 95%CI 0.02, 0.07) and widowed/divorcee respondents (OR 0.72; 95%CI 0.56, 0.82) were less likely to receive a Pap smear test. Educated respondents were more likely to have had a Pap smear test. CONCLUSIONS: The overall prevalence of cervical cancer screening in Malaysia remains low (35.2%). Efforts should be made to strengthen health promotion programs and policies in increasing awareness on the significance of cervical cancer screening. These initiatives should specifically target younger women, single women, and widowed/divorced individuals. The higher cervical screening uptake among rural women should be studied further, and the enabling factors in the rural setup should be emulated in urban areas whenever possible.

Malaysian burden of disease: years of life lost due to premature deaths.

BACKGROUND: In Malaysia, the previous mortality burden has been a significant concern, particularly due to the high prevalence of noncommunicable diseases (NCDs) as the leading cause of death. Estimates of mortality are key indicators for monitoring population health and determining priorities in health policies and health planning. The aim of this study was to estimate the disease burden attributed to 113 major diseases and injuries in Malaysia in 2018 using years of life lost (YLL) method. METHODS: This study included all deaths that occurred in Malaysia in 2018. The YLL was derived by adding the number of deaths from 113 specific diseases and multiplying it by the remaining life expectancy for that age and sex group. Data on life expectancy and mortality were collected from the Department of Statistics Malaysia. RESULTS: In 2018, there were 3.5 million YLL in Malaysia. Group II (NCDs) caused 72.2% of total YLL. Ischaemic heart disease was the leading cause of premature mortality among Malaysians (17.7%), followed by lower respiratory infections (9.7%), road traffic injuries (8.7%), cerebrovascular disease (stroke) (8.0%), and diabetes mellitus (3.9%). CONCLUSIONS: NCDs are a significant health concern in Malaysia and are the primary contributor to the overall burden of disease. These results are important in guiding the national health systems on how to design and implement effective interventions for NCDs, as well as how to prioritise and allocate healthcare resources. Key strategies to consider include implementing health promotion campaigns, adopting integrated care models, and implementing policy and regulatory measures. These approaches aim to enhance health outcomes and the managements of NCDs in Malaysia.

Conditional or unconditional logistic regression for frequency matched case-control design?

Frequency matching is commonly used in epidemiological case control studies to balance the distributions of the matching factors between the case and control groups and to improve the efficiency of case-control designs. Applied researchers have held a common opinion that unconditional logistic regression should be used to analyze frequency matched designs and conditional logistic regression is unnecessary. However, the justification of this view is unclear. To compare the performances of ULR and CLR in terms of simplicity, unbiasedness, and efficiency in a more intuitive way, we viewed frequency matching from the perspective of weighted sampling and derived the outcome models describing how the exposure and matching factors are associated with the outcome in the matched data separately in two scenarios: (1) only categorical variables are used for matching; (2) continuous variables are categorized for matching. In either scenario the derived outcome model is a logit model with stratum-specific intercepts. Correctly specified unconditional logistic regression can be more efficient than conditional logistic regression, particularly when continuous matching factors are used, whereas conditional logistic regression is a more practical approach because it is less dependent on modeling choices.

Comparing statistical methods in assessing the prognostic effect of biomarker variability on time-to-event clinical outcomes.

BACKGROUND: In recent years there is increasing interest in modeling the effect of early longitudinal biomarker data on future time-to-event or other outcomes. Sometimes investigators are also interested in knowing whether the variability of biomarkers is independently predictive of clinical outcomes. This question in most applications is addressed via a two-stage approach where summary statistics such as variance are calculated in the first stage and then used in models as covariates to predict clinical outcome in the second stage. The objective of this study is to compare the relative performance of various methods in estimating the effect of biomarker variability. METHODS: A joint model and 4 different two-stage approaches (naïve, landmark analysis, time-dependent Cox model, and regression calibration) were illustrated using data from a large multi-center randomized phase III trial, the Ocular Hypertension Treatment Study (OHTS), regarding the association between the variability of intraocular pressure (IOP) and the development of primary open-angle glaucoma (POAG). The model performance was also evaluated in terms of bias using simulated data from the joint model of longitudinal IOP and time to POAG. The parameters for simulation were chosen after OHTS data, and the association between longitudinal and survival data was introduced via underlying, unobserved, and error-free parameters including subject-specific variance. RESULTS: In the OHTS data, joint modeling and two-stage methods reached consistent conclusion that IOP variability showed no significant association with the risk of POAG. In the simulated data with no association between IOP variability and time-to-POAG, all the two-stage methods (except the naïve approach) provided a reliable estimation. When a moderate effect of IOP variability on POAG was imposed, all the two-stage methods underestimated the true association as compared with the joint modeling while the model-based two-stage method (regression calibration) resulted in the least bias. CONCLUSION: Regression calibration and joint modelling are the preferred methods in assessing the effect of biomarker variability. Two-stage methods with sample-based measures should be used with caution unless there exists a relatively long series of longitudinal measurements and/or strong effect size (NCT00000125).

The burden of premature mortality among older adults: a population-based study in Malaysia.

BACKGROUND: The populations of many countries-including Malaysia-are rapidly growing older, causing a shift in leading causes of disease and death. In such rapidly ageing populations, it is critical to monitor trends in burden of disease and health of older adults by identifying the leading causes of premature mortality and measuring years of life lost (YLL) to these. The objective of this study, therefore, is to describe the burden (quantified by YLL) associated with major causes of premature mortality among older adults in Malaysia in 2019. METHODS: All deaths that occurred in older adults aged 60 and above in Malaysia in the year 2019 were included in this study. YLL was calculated by summing the number of deaths for the disease category at 5-year age intervals, multiplied by the remaining life expectancy for the specific age and sex group. Both life expectancy and mortality data were obtained from the Department of Statistics Malaysia. RESULTS: In 2019, older adults accounted for 67.4% of total deaths in Malaysia (117,102 out of 173,746). The total number of YLL among older adults in Malaysia in 2019 was estimated at 1.36 million YLL, accounting for 39.6% of the total YLL (3.44 million) lost to all premature deaths in that year. The major causes of premature mortality among older adults were ischaemic heart disease (29.5%) followed by cerebrovascular disease (stroke) (20.8%), lower respiratory infections (15.9%), diabetes mellitus (8.1%) and trachea, bronchus and lung cancers (5.0%). CONCLUSIONS: Non-communicable diseases (NCD) remained the largest contributor to premature mortality among older adults in Malaysia. Implementation of population-level NCD health promotion programmes, screening programmes among high-risk groups and holistic intervention programmes among populations living with NCD are critical in reducing the overall burden of premature mortality.

Ebola Entry Inhibitors Discovered from Maesa perlarius.

Ebola virus disease (EVD), a disease caused by infection with Ebola virus (EBOV), is characterized by hemorrhagic fever and a high case fatality rate. With limited options for the treatment of EVD, anti-Ebola viral therapeutics need to be urgently developed. In this study, over 500 extracts of medicinal plants collected in the Lingnan region were tested against infection with Ebola-virus-pseudotyped particles (EBOVpp), leading to the discovery of Maesa perlarius as an anti-EBOV plant lead. The methanol extract (MPBE) of the stems of this plant showed an inhibitory effect against EBOVpp, with an IC50 value of 0.52 µg/mL, which was confirmed by testing the extract against infectious EBOV in a biosafety level 4 laboratory. The bioassay-guided fractionation of MPBE resulted in three proanthocyanidins (procyanidin B2 (1), procyanidin C1 (2), and epicatechin-(4ß→8)-epicatechin-(4ß→8)-epicatechin-(4ß→8)-epicatechin (3)), along with two flavan-3-ols ((+)-catechin (4) and (-)-epicatechin (5)). The IC50 values of the compounds against pseudovirion-bearing EBOV-GP ranged from 0.83 to 36.0 µM, with 1 as the most potent inhibitor. The anti-EBOV activities of five synthetic derivatives together with six commercially available analogues, including EGCG ((-)-epigallocatechin-3-O-gallate (8)), were further investigated. Molecular docking analysis and binding affinity measurement suggested the EBOV glycoprotein could be a potential molecular target for 1 and its related compounds.

3,4-Seco-Isopimarane Diterpenes from the Twigs and Leaves of Isodon Flavidus.

Three isopimarane diterpenes [fladins B (1), C (2), and D (3)] were isolated from the twigs and leaves of Chinese folk medicine, Isodon flavidus. The chemical structures were determined by the analysis of the comprehensive spectroscopic data, and the absolute configuration was confirmed by X-ray crystallographic analysis. The structures of 1-3 were formed from isopimaranes through the rearrangement of ring A by the bond break at C-3 and C-4 to form a new δ-lactone ring system between C-3 and C-9. This structure type represents the first discovery of a natural isopimarane diterpene with an unusual lactone moiety at C-9 and C-10. In the crystal of 1, molecules are linked to each other by intermolecular O-H···O bonds, forming chains along the b axis. Compounds 1-3 were evaluated for their bioactivities against different diseases. None of these compounds displayed cytotoxic activities against HCT116 and A549 cancer cell lines, antifungal activities against Trichophyton rubrum and T. mentagrophytes, or antiviral activities against HIV entry at 20 µg/mL (62.9-66.7) µM. Compounds 1 and 3 did not show antiviral activities against Ebola entry at 20 µg/mL either; only 2 was found to show an 81% inhibitory effect against Ebola entry activity at 20 µg/mL (66.7 µM). The bioactivity evidence suggested that this type of compound could be a valuable antiviral lead for further structure modification to improve the antiviral potential.

Matched Versus Unmatched Analysis of Matched Case-Control Studies.

Although the need for addressing matching in the analysis of matched case-control studies is well established, debate remains as to the most appropriate analytical method when matching on at least 1 continuous factor. We compared the bias and efficiency of unadjusted and adjusted conditional logistic regression (CLR) and unconditional logistic regression (ULR) in the setting of both exact and nonexact matching. To demonstrate that case-control matching distorts the association between the matching variables and the outcome in the matched sample relative to the target population, we derived the logit model for the matched case-control sample under exact matching. We conducted simulations to validate our theoretical conclusions and to explore different ways of adjusting for the matching variables in CLR and ULR to reduce biases. When matching is exact, CLR is unbiased in all settings. When matching is not exact, unadjusted CLR tends to be biased, and this bias increases with increasing matching caliper size. Spline smoothing of the matching variables in CLR can alleviate biases. Regardless of exact or nonexact matching, adjusted ULR is generally biased unless the functional form of the matched factors is modeled correctly. The validity of adjusted ULR is vulnerable to model specification error. CLR should remain the primary analytical approach.

Acute kidney injury post-heart transplant: An analysis of peri-operative risk factors.

Acute kidney injury is a common complication following heart transplantation, and the factors contributing to acute kidney injury are not well understood. We conducted a retrospective cohort study evaluating patients who underwent heart transplantation between 2009 and 2016 at a single institution. The primary endpoint was incidence of acute kidney injury as defined by Kidney Disease Improving Global Outcomes criteria. Secondary endpoints included 30-day hospital readmission, 30-day mortality, and 1-year mortality. A total of 228 heart transplant patients were included in the study for analysis. In total, 145 (64%) developed acute kidney injury, where 43 (30%) were classified as stage I, 28 (19%) as stage II, and 74 (51%) as stage III. Risk factors found to be associated with the presence of acute kidney injury included increased use of vasopressors and inotropes post-transplant. Protective factors included cardiopulmonary bypass time <170 min. Acute kidney injury was found to be associated with increased 30-day and 1-year mortality.

Statistical analysis of two arm randomized pre-post designs with one post-treatment measurement.

BACKGROUND: Randomized pre-post designs, with outcomes measured at baseline and after treatment, have been commonly used to compare the clinical effectiveness of two competing treatments. There are vast, but often conflicting, amount of information in current literature about the best analytic methods for pre-post designs. It is challenging for applied researchers to make an informed choice. METHODS: We discuss six methods commonly used in literature: one way analysis of variance ("ANOVA"), analysis of covariance main effect and interaction models on the post-treatment score ("ANCOVAI" and "ANCOVAII"), ANOVA on the change score between the baseline and post-treatment scores ("ANOVA-Change"), repeated measures ("RM") and constrained repeated measures ("cRM") models on the baseline and post-treatment scores as joint outcomes. We review a number of study endpoints in randomized pre-post designs and identify the mean difference in the post-treatment score as the common treatment effect that all six methods target. We delineate the underlying differences and connections between these competing methods in homogeneous and heterogeneous study populations. RESULTS: ANCOVA and cRM outperform other alternative methods because their treatment effect estimators have the smallest variances. cRM has comparable performance to ANCOVAI in the homogeneous scenario and to ANCOVAII in the heterogeneous scenario. In spite of that, ANCOVA has several advantages over cRM: i) the baseline score is adjusted as covariate because it is not an outcome by definition; ii) it is very convenient to incorporate other baseline variables and easy to handle complex heteroscedasticity patterns in a linear regression framework. CONCLUSIONS: ANCOVA is a simple and the most efficient approach for analyzing pre-post randomized designs.

Assessing the Feasibility of an Online Training Designed to Enable Community Health Workers to Deliver a Comprehensive, Evidence-Based Weight Loss Intervention for Rural African Americans of Faith.

Obesity is a critical modifiable risk factor in cancer prevention, control, and survivorship. Comprehensive weight loss interventions (e.g., Diabetes Prevention Program (DPP)) have been recommended by governmental agencies to treat obesity. However, their high implementation costs limit their reach, especially in underserved African American (AA) communities. Community health workers (CHWs) or trusted community members can help increase access to obesity interventions in underserved regions facing provider shortages. CHW-led interventions have increased weight loss. However, in-person CHW training can be costly to deliver and often requires extensive travel to implement. Web-based trainings have become common to increase reach at reduced cost. However, the feasibility of an online CHW training to deliver the DPP in AAs is unknown. The feasibility of an online CHW training to deliver the DPP adapted for AAs was assessed. The online training was compared to an in-person DPP training with established effectiveness. CHW effectiveness and satisfaction were assessed at baseline and 6 weeks. Nineteen participants (in-person n = 10; online n = 9) were recruited. At post-training, all scored higher than the 80% on a knowledge test required to deliver the intervention. All participants reported high levels of training satisfaction (88.9% of online participants and 90% of in-person participants rated the training as at least 6 on a 1-7 scale) and comfort to complete intervention tasks (78% of online participants and 60% of in-person participants scored at least 6 on a 1-7 scale). There were no significant differences in outcomes by arm. An online CHW training to deliver the DPP adapted for AAs faith communities produced comparable effectiveness and satisfaction to an evidence-based in-person CHW training. Further research is needed to assess the cost-effectiveness of different CHW training modalities to reduce obesity.

Prevalence of Ototoxicity Following Hematopoietic Stem Cell Transplantation in Pediatric Patients.

The use of hematopoietic stem cell transplantation (HSCT) is increasing for a variety of diseases. Ototoxicity from this procedure has not been extensively studied. A retrospective chart review examined 275 patients from this institution who underwent HSCT between January 1, 2007, and April 30, 2017. Data extracted included therapy before HSCT and the subsequent course of transplantation. Evaluable patients had complete medical records and interpretable audiograms available. Ototoxicity constituted significant threshold changes from baseline or changes in International Society of Pediatric Oncology/Boston Ototoxicity Grading Scale (SIOP) grade comparing audiogram results just before HSCT with those following the transplantation procedure. A total of 147 patients were evaluable. Ototoxicity was observed in 10.2% of the patients. Higher SIOP grade before HSCT was significantly associated with a higher risk of post-transplantation ototoxicity (P < .01). Previous cisplatin treatment (P < .0001), but not carboplatin or radiation treatment, was also associated with ototoxicity. Patients with a solid tumor or brain tumor (P < .0001) and those who received an autologous transplant (P = .0002) were also at increased risk. No post-transplantation event was significantly associated with ototoxicity. Ototoxicity affects a significant percentage of patients undergoing HSCT, and careful monitoring is needed to identify patients impacted by this procedure.

A Pilot Study of Lenalidomide Maintenance Therapy after Autologous Transplantation in Relapsed or Refractory Classical Hodgkin Lymphoma.

For patients with relapsed or refractory classical Hodgkin lymphoma (cHL), salvage chemotherapy followed by consolidation with autologous stem cell transplant (ASCT) remains the standard of care. Even with this aggressive treatment strategy, 5-year progression-free survival is ≤50%, and there remains interest in maintenance strategies to improve long-term disease-free survival. Lenalidomide is an immunomodulatory agent with demonstrated activity in multiple subtypes of lymphoma including cHL, and has also been shown to improve both progression-free and overall survival as maintenance therapy after ASCT in multiple myeloma. This multicenter study evaluated maintenance lenalidomide after ASCT for patients with cHL. Patients were enrolled 60 to 90 days post-transplant and received oral lenalidomide on days 1 to 28 of 28-day cycles for a maximum of 18 cycles. Lenalidomide was started at 15 mg daily and increased to maximum of 25 mg daily if tolerated. The primary objective of this study was to assess the feasibility of this regimen, with a goal <30% rate of discontinuation at or before cycle 12 for drug-related reasons. Twenty-seven patients were enrolled and 26 received at least 1 dose of lenalidomide. With a median follow-up of 51.3 months (range, 12.2 to 76.2 months), 23 of 26 patients were alive. Median event-free survival was 9.4 months and median progression-free survival had not been reached, with 17 of 26 patients (65.4%) remaining in remission at last follow-up. Excluding 4 patients who discontinued therapy for progression and 2 who discontinued due to noncompliance, the discontinuation rate at or before cycle 12 was 52%. Treatment was complicated by a high frequency of hematologic adverse events, with 15 patients (58%) experiencing grade 3 to 4 hematologic toxicity and 5 (19%) experiencing grade 4 hematologic toxicity. We conclude that the regimen of maintenance lenalidomide explored in this study is not feasible for patients with cHL immediately following ASCT. An alternative lenalidomide dose or schedule may be better tolerated following ASCT for patients with relapsed or refractory cHL.

Matched or unmatched analyses with propensity-score-matched data?

Propensity-score matching has been used widely in observational studies to balance confounders across treatment groups. However, whether matched-pairs analyses should be used as a primary approach is still in debate. We compared the statistical power and type 1 error rate for four commonly used methods of analyzing propensity-score-matched samples with continuous outcomes: (1) an unadjusted mixed-effects model, (2) an unadjusted generalized estimating method, (3) simple linear regression, and (4) multiple linear regression. Multiple linear regression had the highest statistical power among the four competing methods. We also found that the degree of intraclass correlation within matched pairs depends on the dissimilarity between the coefficient vectors of confounders in the outcome and treatment models. Multiple linear regression is superior to the unadjusted matched-pairs analyses for propensity-score-matched data.

Association of platelet function with depression and its treatment with sertraline in patients with chronic kidney disease: analysis of a randomized trial.

BACKGROUND: Major Depressive Disorder (MDD) can lead to adverse cardiovascular outcomes in patients with chronic kidney disease (CKD). Although one of the proposed mechanisms is heightened platelet activation, effects of MDD and its treatment with a selective serotonin reuptake inhibitor (SSRI) on platelet function in patients with CKD remain unclear. METHODS: In a pre-specified analysis, changes from baseline to 12 weeks in whole blood platelet aggregation (WBPA) and plasma levels of E-selectin and P-selectin on treatment with sertraline vs. placebo were investigated in 175 patients with CKD (estimated glomerular filtration rate [eGFR] < 60 ml/min/1.73m2) and MDD (MDD+/CKD+) in a randomized, double-blind trial. Correlations between severity of depressive symptoms and platelet function were also analyzed. In order to investigate whether differences in platelet function were due to presence of CKD or MDD, we compared a subgroup of 49 MDD+/CKD+ patients with eGFR < 30 ml/min/1.73m2 to 43 non-depressed CKD controls (28 CKD with eGFR < 30 ml/min/1.73m2 [MDD-/CKD+] and 15 individuals with eGFR ≥90 ml/min/1.73m2 [MDD-/CKD-]. RESULTS: In MDD+/CKD+ individuals, there were no significant correlations between severity of depressive symptoms and platelet function, and no significant changes in platelet function after 12 weeks of treatment with sertraline vs. placebo. There were no significant differences in platelet function among MDD+/CKD+ patients and controls without MDD except in WBPA to 10 µM ADP (P = 0.03). WBPA to ADP was lower in the MDD-/CKD- group (8.0 Ω [5.0 Ω, 11.0 Ω]) as compared to the MDD-/CKD+ group (12.5 Ω [8.0 Ω, 14.5 Ω]), P = 0.01, and the MDD+/CKD+ group (11.0 Ω [8.0 Ω, 15.0 Ω]), P < 0.01. CONCLUSIONS: Heightened ADP-induced platelet aggregability was observed in CKD patients compared to controls with normal kidney function, regardless of presence of comorbid MDD, and treatment with sertraline did not affect platelet function. These findings suggest that increased platelet activation may not be a major contributory underlying mechanism by which depression may lead to worse cardiovascular outcomes in patients with CKD. Future studies should include positive MDD controls without CKD to confirm our findings. TRIAL REGISTRATION: ClinicalTrials.gov identifier numbers: CAST Study: NCT00946998 (Recruitment Status: Completed. First Posted: July 27, 2009. Results First Posted: January 30, 2018). WiCKDonASA Study: NCT01768637 (Recruitment Status: Completed. First Posted: January 15, 2013. Results First Posted: April 19, 2019).

A general approach to evaluating the bias of 2-stage instrumental variable estimators.

Unmeasured confounding is a common concern when researchers attempt to estimate a treatment effect using observational data or randomized studies with nonperfect compliance. To address this concern, instrumental variable methods, such as 2-stage predictor substitution (2SPS) and 2-stage residual inclusion (2SRI), have been widely adopted. In many clinical studies of binary and survival outcomes, 2SRI has been accepted as the method of choice over 2SPS, but a compelling theoretical rationale has not been postulated. We evaluate the bias and consistency in estimating the conditional treatment effect for both 2SPS and 2SRI when the outcome is binary, count, or time to event. We demonstrate analytically that the bias in 2SPS and 2SRI estimators can be reframed to mirror the problem of omitted variables in nonlinear models and that there is a direct relationship with the collapsibility of effect measures. In contrast to conclusions made by previous studies (Terza et al, 2008), we demonstrate that the consistency of 2SRI estimators only holds under the following conditions: (1) when the null hypothesis is true; (2) when the outcome model is collapsible; or (3) when estimating the nonnull causal effect from Cox or logistic regression models, the strong and unrealistic assumption that the effect of the unmeasured covariates on the treatment is proportional to their effect on the outcome needs to hold. We propose a novel dissimilarity metric to provide an intuitive explanation of the bias of 2SRI estimators in noncollapsible models and demonstrate that with increasing dissimilarity between the effects of the unmeasured covariates on the treatment versus outcome, the bias of 2SRI increases in magnitude.