Immeasurable Time Bias Research Articles

Statins and Mortality in COPD: A Methodological Review of Observational Studies

Randomized controlled trials and observational studies have reported conflicting results on the potential beneficial effects of statins on mortality in patients with chronic obstructive pulmonary disease (COPD). We performed a systematic search of the literature to review all observational studies reporting relative risks of death with statin use in COPD, focusing on potential sources of bias. We identified 15 observational studies, out of 2835, of which 12 were affected by time-related and other biases and the remaining 3 by confounding bias. All 15 studies were also subject to confounding bias due to lack of adjustment for important COPD-related factors. The risk of death associated with statin use was reduced across all 15 studies (pooled relative risk (PRR) 0.66; 95% CI: 0.59-0.74). The reduction was observed in 7 studies with immortal time bias (PRR 0.62; 95%: 0.53-0.72), two with collider-stratification bias (PRR 0.60; 95% CI: 0.45-0.80), one with time-window bias (RR 0.61; 95% CI: 0.38-0.98), one with immeasurable time bias (RR 0.50; 95% CI: 0.40-0.62), and one with exposure misclassification (RR 0.86; 95% CI: 0.72-1.03). The three studies that avoided these biases were, however, affected by confounding bias resulting in a PRR of 0.77 (95% CI: 0.61-0.98). In conclusion, the observational studies investigating statin use and mortality in COPD are affected by major biases, many of which can result in spurious protective effects. Well-designed observational studies that carefully emulate randomized trials are needed to resolve this uncertainty regarding the potential beneficial benefits of statins on mortality in patients with COPD.

Validating an approach to overcome the immeasurable time bias in cohort studies: a real-world example and Monte Carlo simulation study.

Immeasurable time bias arises from the lack of in-hospital medication information. It has been suggested that time-varying adjustment for hospitalization may minimize this potential bias. However, whereas we examined this issue in one case study, there remains a need to assess the validity of this approach in other settings. Using a Monte Carlo simulation, we generated synthetic immeasurable time-varying hospitalization-related factors of duration, frequency and timing. Nine scenarios were created by combining three frequency scenarios and three duration scenarios, where the empirical cohort distribution of hospitalization was used to simulate the 'timing'. We used Korea's healthcare database and a case example of β-blocker use and mortality among patients with heart failure. We estimated the gold-standard hazard ratio (HR) with 95% CI using inpatient and outpatient drug data, and that of the pseudo-outpatient setting using outpatient data only. We assessed the validity of adjusting for time-varying hospitalization in nine different scenarios, using relative bias, confidence limit ratio (CLR) and mean squared error (MSE) compared with the empirical gold-standard estimate across bootstrap resamples. With the real-world gold standard (HR 0.73; 95% CI 0.67-0.80) as the reference estimate, adjusting for time-varying hospitalization (0.71; 0.63-0.80) effectively reduced the immeasurable time bias and had the following performance metrics across the nine scenarios: relative bias (range: -7.08% to 0.61%), CLR (1.28 to 1.36) and MSE (0.0005 to 0.0031). The approach of adjusting for time-varying hospitalization consistently reduced the immeasurable time bias in Monte Carlo simulated data.

Bleeding risk and mortality according to antithrombotic agents’ exposure in cancer-related stroke patients: nationwide population-based cohort study in South Korea

BackgroundIschemic stroke with active cancer is thought to have a unique mechanism compared to conventional stroke etiologies. There is no gold standard guideline for secondary prevention in patients with cancer-related stroke, hence, adequate type of antithrombotic agent for treatment is controversial.MethodsSubjects who were enrolled in National Health Insurance System Customized Research data during the period between 2010 and 2015 were observed until 2019. Subject diagnosed with ischemic stroke within six months before and 12 months after a cancer diagnosis was defined as cancer-related stroke patient. To solve immeasurable time bias, the drug exposure evaluation was divided into daily units, and each person-day was classified as four groups: antiplatelet, anticoagulant, both types, and unexposed to antithrombotic drugs. To investigate bleeding risk and mortality, Cox proportional hazards regression model with time-dependent covariates were used.ResultsTwo thousand two hundred eighty-five subjects with cancer-related stroke were followed and analyzed. A group with anticoagulation showed high estimated hazard ratios (HRs) of all bleeding events compared to a group with antiplatelet (major bleeding HR, 1.35; 95% confidence interval [CI], 1.20–1.52; p < 0.001). And the result was also similar in the combination group (major bleeding HR, 1.54; 95% CI, 1.13–2.09; p = 0.006). The combination group also showed increased mortality HR compared to antiplatelet group (HR, 1.72; 95% CI, 1.47–2.00; p < 0.001).ConclusionsBleeding risk increased in the anticoagulant-exposed group compared to antiplatelet-exposed group in cancer-related stroke patients. Thus, this result should be considered when selecting a secondary prevention drug.

Analytical Approaches to Reduce Selection Bias in As-Treated Analyses with Missing In-Hospital Drug Information.

While much attention has focused on immeasurable time bias as a potential exposure misclassification bias, it may also result in potential selection bias in cohort studies using an as-treated (or per protocol) exposure definition in which patients are censored upon treatment discontinuation. We examined analytical approaches to minimise informative censoring due to the absence of in-hospital drug data using a case study of β-blocker use and mortality in heart failure. We conducted a cohort study using Korea's healthcare database, including inpatient and outpatient drug data. Using an as-treated exposure definition, patients were followed up until death, β-blocker discontinuation (in the exposed), β-blocker initiation (in the unexposed), or end of study period. In 'complete prescription' analysis using inpatient and outpatient drug data, we estimated hazard ratios (HR) and 95% confidence intervals (CI) using a Cox proportional hazard model. In outpatient drug-based analyses, we attempted to reduce the bias using stabilised inverse probability weighting (IPW) for treatment crossovers, hospitalisation, and all artificial censorings. An HR of 0.89 (95% CI 0.74-1.07) for β-blocker use versus non-use for all-cause mortality was found in 'complete prescription' analysis. Benefits were exaggerated when follow-up was assessed using outpatient drug data only (HR 0.71; 95% CI 0.57-0.89). Weighting by stabilised IPW for treatment crossovers and hospitalisation reduced the bias. When using an as-treated exposure definition, missing in-hospital drug data induced selection bias in our case study. Using IPW for censoring mitigated bias from the hospitalisation-induced censorings.

Effectiveness of Allopurinol in Reducing Mortality: Time-Related Biases in Observational Studies.

The treatment of gout with allopurinol is effective at reducing urate levels and the frequency of flares. Several observational studies have shown important reductions in mortality with allopurinol use, with wide variations in results. We undertook this review to assess the extent of bias in these studies, particularly time-related biases such as immortal time bias. We searched the literature to identify all observational studies describing the effect of allopurinol use versus nonuse on all-cause mortality. We identified 12 observational studies, of which 3 were affected by immortal time bias and 3 by immeasurable time bias, while the remaining 6 studies avoided these time-related biases. Reductions in all-cause mortality with allopurinol use were observed among the studies with immortal time bias, with a pooled hazard ratio (HR) of death associated with allopurinol of 0.71 (95% confidence interval [95% CI] 0.50-1.01), as well as in those with immeasurable time bias (pooled HR 0.62 [95% CI 0.56-0.67]). The 6 studies that avoided these biases demonstrated a null effect of allopurinol on mortality (pooled HR 0.99 [95% CI 0.87-1.11]), though the lack of an analysis based on treatment adherence may have attenuated the effect. Observational studies are important to provide real-world data on medication effects. The observational studies showing significantly decreased mortality with allopurinol treatment cannot be used as evidence, however, mainly due to time-related biases that tend to greatly exaggerate the potential benefit of treatments. The ALL-HEART randomized trial, which is currently underway and evaluates the effect of adding allopurinol to usual care (compared to no added treatment), will provide reliable evidence on mortality.

Immeasurable Time Bias in Self-controlled Designs: Case-crossover, Case-time-control, and Case-case-time-control Analyses.

Impact of immeasurable time bias (IMTB) is yet to be examined in self-controlled designs. We conducted case-crossover, case-time-control, and case-case-time-control analyses using Korea's healthcare database. Two empirical examples among elderly patients were used: 1) benzodiazepines-hip fracture; 2) benzodiazepines-mortality. For cases, the date of hip fracture diagnosis or death was defined as the index date, and the inherited date of their matched cases for controls or future cases. Exposure was assessed in the 1-30 day (hazard) and 61-90 day (control) windows preceding the index date. A non-missing exposure setting included in- and outpatient prescriptions and the pseudo-outpatient setting included only the outpatients. Conditional logistic regression was done to estimate odds ratios (ORs) with 95% confidence intervals (CIs), where the relative difference in OR among the two settings was calculated to quantify the IMTB. The IMTB had negligible impacts in the hip fracture example in the case-crossover (non-missing exposure setting OR 1.27; 95% CI, 1.12-1.44; pseudo-outpatient setting OR 1.21; 95% CI, 1.06-1.39; magnitude 0.05), case-time-control (OR 1.18; 95% CI, 0.98-1.44; OR 1.13; 95% CI, 0.92-1.38; 0.04, respectively), and case-case-time-control analyses (OR 0.99; 95% CI, 0.80-1.23; OR 0.94; 95% CI, 0.75-1.18; 0.05, respectively). In the mortality example, IMTB had significant impacts in the case-crossover (non-missing exposure setting OR 1.44; 95% CI, 1.36-1.52; pseudo-outpatient setting OR 0.72; 95% CI, 0.67-0.78; magnitude 1.00), case-time-control (OR 1.38; 95% CI, 1.26-1.51; OR 0.68; 95% CI, 0.61-0.76; 1.03, respectively), and case-case-time-control analyses (OR 1.27; 95% CI, 1.15-1.40; OR 0.62; 95% CI, 0.55-0.69; 1.05, respectively). Although IMTB had negligible impacts on the drug's effect on acute events, as these are unlikely to be accompanied with hospitalizations, it negatively biased the drug's effect on mortality, an outcome with prodromal phases, in the three self-controlled designs.

Analytical approaches to minimizing immeasurable time bias in cohort studies.

Immeasurable time bias exaggerates drug benefits in pharmacoepidemiological studies due to exposure misclassification arising from the inability to measure in-hospital medications in many health care databases. To compare the ability of different methodological approaches to minimize immeasurable time bias, we conducted a cohort study of β-blocker use and all-cause mortality among patients with heart failure (HF), using a nationwide health care database which contains both in- and outpatient prescriptions. In our gold-standard analysis, we assessed exposure using a time-varying approach involving both in- and outpatient prescriptions. Cox proportional hazard models were used to estimate hazard ratios (HRs) with 95% confidence intervals (CIs) of mortality, with exposure to β-blockers defined as a time-varying variable. To estimate the magnitude of the immeasurable time bias, we repeated the analyses using outpatient prescriptions only and compared 10 approaches to minimize the bias, which are categorized as restriction, adjustment, assumption and weighting. The HR for β-blocker use versus non-use was 0.76 (95% CI: 0.71 to 0.80) in our gold-standard analysis. When exposure assessment was restricted to outpatient prescriptions only, β-blocker use was substantially more protective (HR 0.43, 95% CI: 0.40 to 0.46). Of the 10 approaches examined, adjusting for hospitalization as a time-varying variable successfully minimized the bias (HR 0.75, 95% CI: 0.68 to 0.82). The immeasurable time bias can result in substantial bias in pharmacoepidemiological studies. Time-varying adjustment for hospitalization appears to reduce the immeasurable time bias in the absence of inpatient medication data.

New methodological approaches were able to effectively reduce immeasurable time bias in case-only designs

ObjectivesThe objective of this study was to assess approaches to reduce immeasurable time bias in case-crossover (CCO), case-time-control (CTC), and case-case-time-control (CCTC) designs. Study Design and SettingWe used Korea's health care database that has inpatient and outpatient prescriptions and an empirical example of benzodiazepines and mortality among the elderly. We defined our unbiased exposure setting using all prescriptions and a pseudo-outpatient setting using outpatient records only. In the pseudo-outpatient setting, we assessed 10 approaches of restricting, adjusting, stratifying, or weighting on hospitalization-related factors. We conducted conditional logistic regression to estimate odds ratio (OR) with 95% confidence intervals (CI), where an approach was considered effective when its OR was within the unbiased exposure setting OR's 95% CI. ResultsImmeasurable time bias negatively biased the unbiased exposure setting's OR in all three case-only designs, overestimating the protective effect of benzodiazepines on mortality. Of the 10 approaches examined, stratifying the proportion of hospitalized time in 0.01 intervals most effectively repaired the bias in the CCO (OR 1.25, 95% CI 1.10–1.43) and CTC analyses (1.11, 0.95–1.30); no approach was effective in the CCTC analysis. ConclusionStratifying the proportion of hospitalized time in 0.01 intervals best approximated the unbiased exposure setting estimate by overcoming the significant impact of immeasurable time bias in CCO and CTC designs.

An empirical assessment of immeasurable time bias in the setting of nested case-control studies: Statins and all-cause mortality among patients with heart failure.

Immeasurable time bias exaggerates drug benefits in pharmacoepidemiologic studies due to exposure misclassification that occurs due to the lack of inpatient drug data in many healthcare databases. To estimate the magnitude of immeasurable time bias and assess potential approaches to minimize it, we conducted a nested case-control study of statin use and mortality among heart failure patients using the South Korean nationwide healthcare database, which contains both inpatient and outpatient medication data. Using both inpatient and outpatient medication data to define the gold standard exposure definition, we assessed 10 different analytical methods in which exposure was defined using outpatient medication data only. We compared different methodological approaches to reduce immeasurable time bias: restricting to nonhospitalized patients, adjusting for hospitalization, weighting by either measurable time (nonhospitalized time during 90-d period) or outpatient time, and computing the odds ratios (ORs) using 90-day cumulative probability of exposure produced by the Kaplan-Meier product-limit estimator for cases and controls. The three approaches that most closely approximated the gold standard (hazard ratio [HR] 1.20; 95% confidence interval [CI], 1.05-1.37) were weighting by either measurable (HR 1.09; 95% CI, 0.92-1.28) or outpatient time (HR 1.14; 95% CI, 0.96-1.34) in the unexposed or by estimating the 90-day exposure probability (HR 1.31; 95% CI, 1.11-1.51). The use of one of these three methods may be suggested as an approach to minimize immeasurable time bias in nested case-control studies.

Observational Studies of Inhaled Corticosteroid Effectiveness in COPD: Lessons Learned

Randomized controlled trials at times investigate findings suggested by observational studies. For example, the Towards a Revolution in COPD Health (TORCH) trial, which did not show a mortality reduction with inhaled corticosteroids (ICS) in COPD, was motivated by some observational studies that suggested considerable reductions in mortality with these drugs. Reasons for these discrepancies are unclear. The literature was searched to identify all observational studies, including cohort and case-control studies, investigating the effectiveness of ICS on major outcomes in patients with COPD; these outcomes included death and hospitalization for COPD. A total of 21 studies were identified. Nine studies were affected by immortal time bias, five by immeasurable time bias, and seven by the "asthma factor" bias; some studies were affected by more than one bias. These studies found important reductions in the rates of major COPD outcomes with ICS use, with pooled rate ratios of 0.71 (95%CI, 0.67-0.76), 0.76 (95%CI, 0.70-0.83), and 0.79 (95%CI, 0.73-0.87), respectively, for the three sources of bias. In contrast, the five studies unaffected by these major biases did not find an association (pooled rate ratio, 1.02 [95%CI, 0.88-1.17]). Observational studies are important to provide evidence from real-world data on medication effects. However, appropriate study design and analysis are essential to avoid biases and ensure randomized trials with greater chances of success. The observational studies suggesting potential beneficial effects of nonrespiratory drugs to treat COPD, such as statins and beta-blockers, will also need careful review before long and expensive randomized trials are conducted.

Immeasurable time bias due to hospitalization in medico-administrative databases: which impact for pharmacoepidemiological studies?

Drugs administered to hospitalized patients are not available within almost all health insurance databases. However, this unobservable exposure time bias is very rarely taken into account in pharmacoepidemiology. The objective was to model unobservable periods due to hospitalization and to assess their impact on risk estimates in the context of the association between benzodiazepines and mortality. A cohort study was identified using the General Sample of Beneficiaries in France for the period 2006-2012. Benzodiazepines incident users were matched to incident users of antidepressants/non-benzodiazepine sedatives and to controls (non-users) according to age and gender. All-cause mortality at 1year (Cox regression model) was studied using time-dependent variables (exposed/unexposed or under two hypotheses, inpatients are exposed or inpatients are unexposed), complemented with a multistate model based on observable/unobservable/death status. In each group, 57242 patients were included. All-cause mortality was significantly higher among those exposed to benzodiazepines (adjusted hazard ratio (aHR)=1.17, 95% confidence interval (CI)=1.04, 1.32), as compared with controls. Use of benzodiazepines exposure as a time-dependent variable resulted in significant results after adjustment (aHR=1.45, 95%CI=1.16, 1.80). When inpatients were considered as unexposed, all-cause mortality was not significantly increased (aHR=0.85, 95%CI=0.76, 1.10), but was significantly augmented when inpatients were considered as exposed (aHR=2.93, 95%CI=2.46, 3.48). This study highlights the need to take account of the impact of unobservable exposure periods on risk estimates. Copyright © 2017 John Wiley & Sons, Ltd.

Effects of Cardiovascular Drugs on Mortality in Severe Chronic Obstructive Pulmonary Disease

Cardiovascular drugs may improve survival in chronic obstructive pulmonary disease (COPD). However, previous studies did not account for major sources of bias, and drug effects have not been evaluated in severe COPD. To estimate the time-dependent effects of cardiovascular drugs on survival in oxygen-dependent COPD, accounting for immortal and immeasurable time bias. Prospective national study of patients starting long-term oxygen therapy for COPD in Sweden between 1 October 2005 and 30 June 2009. Effects on mortality were estimated using extended Cox regression adjusted for age, sex, PaO2, PaCO2, World Health Organization performance status, body mass index, comorbidity, and concomitant medications. Immortal and immeasurable time bias was addressed by analyzing all medications as time-dependent variables and accounting for hospitalized time, respectively. Time-dependent effects of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, antiplatelet drugs, β-blockers, and statins on all-cause mortality were measured. Of the 2,249 included patients, 1,129 (50%) died under observation. No patient was lost to follow-up. The adjusted time-dependent model was compatible with reduced mortality for antiplatelet drugs (hazard ratio [HR], 0.86; 95% CI, 0.75-0.99; P = 0.030) and trends for angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (HR, 0.90; 95% CI, 0.79-1.04; P = 0.166) and statins (HR, 0.86; 95% CI, 0.72-1.03; P = 0.105), whereas β-blockers increased mortality (HR, 1.19; 95% CI, 1.04-1.37; P = 0.010). This study supports that antiplatelet drugs improve survival and β-blockers decrease survival in oxygen-dependent COPD.

Immeasurable Time Bias in Observational Studies of Drug Effects on Mortality

Observational studies suggesting that some drugs are effective at reducing mortality may have been subject to "immeasurable time bias" arising from the unidentified presence of hospitalizations when defining drug exposure with computerized health databases. The author illustrates the bias using a case-control study of 1,313 deaths and 1,313 controls selected from a cohort of 2,049 patients with chronic obstructive pulmonary disease from Saskatchewan, Canada, identified from 1990 and followed up through 1999. Different approaches were used to estimate the rate ratio of death associated with inhaled corticosteroid exposure, defined by a prescription dispensed in the 30-day period prior to the index date. More cases had been hospitalized during the 30-day exposure period (72%) than controls (26%), with lower durations of stay for cases who received an inhaled corticosteroid prescription (9.9 vs.16.2 days), thus introducing variations in measurable exposure times. The raw analysis that did not consider hospitalization found a rate ratio of 0.60 (95% confidence interval (CI): 0.50, 0.73). Alternatively, analyses accounting for variations in measurable times resulted in a rate ratio of 0.93 (95% CI: 0.76, 1.14) when weighted by measurable time, while use of the Kaplan-Meier estimator of the 30-day cumulative incidence of exposure found a rate ratio of 1.35 (95% CI: 1.14, 1.60). In conclusion, immeasurable time bias may be present in several observational database studies suggesting that certain drugs are effective at reducing mortality.