Evaluating the Validity of an Instrumental Variable Study of Neuroleptics

Comparative Analysis of Instrumental Variables on the Assignment of Buprenorphine/Naloxone or Methadone for the Treatment of Opioid Use Disorder

BACKGROUND

Instrumental variable (IV) analysis provides an alternative set of identification assumptions in the presence of uncontrolled confounding when attempting to estimate causal effects. Our objective was to evaluate the suitability of measures of prescriber preference and calendar time as potential IVs to evaluate the comparative effectiveness of buprenorphine/naloxone versus methadone for treatment of opioid use disorder (OUD).

METHODS

Using linked population-level health administrative data, we constructed five IVs: prescribing preference at the individual, facility, and region levels (continuous and categorical variables), calendar time, and a binary prescriber's preference IV in analyzing the treatment assignment-treatment discontinuation association using both incident-user and prevalent-new-user designs. Using published guidelines, we assessed and compared each IV according to the four assumptions for IVs, employing both empirical assessment and content expertise. We evaluated the robustness of results using sensitivity analyses.

RESULTS

The study sample included 35,904 incident users (43.3% on buprenorphine/naloxone) initiated on opioid agonist treatment by 1585 prescribers during the study period. While all candidate IVs were strong (A1) according to conventional criteria, by expert opinion, we found no evidence against assumptions of exclusion (A2), independence (A3), monotonicity (A4a), and homogeneity (A4b) for prescribing preference-based IV. Some criteria were violated for the calendar time-based IV. We determined that preference in provider-level prescribing, measured on a continuous scale, was the most suitable IV for comparative effectiveness of buprenorphine/naloxone and methadone for the treatment of OUD.

CONCLUSIONS

Our results suggest that prescriber's preference measures are suitable IVs in comparative effectiveness studies of treatment for OUD.

Intravascular imaging during percutaneous coronary intervention: temporal trends and clinical outcomes in the USA

AIMS

Prior trials have demonstrated that intravascular imaging (IVI)-guided percutaneous coronary intervention (PCI) results in less frequent target lesion revascularization and major adverse cardiovascular events (MACEs) compared with standard angiographic guidance. The uptake and associated outcomes of IVI-guided PCI in contemporary clinical practice in the USA remain unclear. Accordingly, temporal trends and comparative outcomes of IVI-guided PCI relative to PCI with angiographic guidance alone were examined in a broad, unselected population of Medicare beneficiaries.

METHODS AND RESULTS

Retrospective cohort study of Medicare beneficiary data from 1 January 2013, through 31 December 2019 to evaluate temporal trends and comparative outcomes of IVI-guided PCI as compared with PCI with angiography guidance alone in both the inpatient and outpatient settings. The primary outcomes were 1 year mortality and MACE, defined as the composite of death, myocardial infarction (MI), repeat PCI, or coronary artery bypass graft surgery. Secondary outcomes were MI or repeat PCI at 1 year. Multivariable Cox regression was used to estimate the adjusted association between IVI guidance and outcomes. Falsification endpoints (hospitalized pneumonia and hip fracture) were used to assess for potential unmeasured confounding. The study population included 1 189 470 patients undergoing PCI (38.0% female, 89.8% White, 65.1% with MI). Overall, IVI was used in 10.5% of the PCIs, increasing from 9.5% in 2013% to 15.4% in 2019. Operator IVI use was variable, with the median operator use of IVI 3.92% (interquartile range 0.36%-12.82%). IVI use during PCI was associated with lower adjusted rates of 1 year mortality [adjusted hazard ratio (aHR) 0.96, 95% confidence interval (CI) 0.94-0.98], MI (aHR 0.97, 95% CI 0.95-0.99), repeat PCI (aHR 0.74, 95% CI 0.73-0.75), and MACE (aHR 0.85, 95% CI 0.84-0.86). There was no association with the falsification endpoint of hospitalized pneumonia (aHR 1.02, 95% CI 0.99-1.04) or hip fracture (aHR 1.02, 95% CI 0.94-1.10).

CONCLUSION

Among Medicare beneficiaries undergoing PCI, use of IVI has increased over the previous decade but remains relatively infrequent. IVI-guided PCI was associated with lower risk-adjusted mortality, acute MI, repeat PCI, and MACE.

Does continuity of care improve patient satisfaction? An instrumental variable approach

Numerous studies have investigated the relationship between continuity of care (COC) and patient satisfaction. However, COC and patient satisfaction were measured simultaneously; therefore, the direction of causality remains understudied. This study examined the effect of COC on the patient satisfaction of elderly individuals using an instrumental variable (IV) approach. Nationwide survey data acquired using a face-to-face interview were used to measure the patient-reported COC experiences of 1,715 participants. We applied an ordered logit model controlled for observed patient characteristics and a two-stage residual inclusion (2SRI) ordered logit model that accounted for unobserved confounding factors. Patient-perceived COC importance was used as an IV for patient-reported COC. The ordered logit models indicated that patients with high or intermediate patient-reported COC scores were more likely to perceive more patient satisfaction than those with low COC scores. Using the patient-perceived COC importance as an IV, we examined a strong significant association between the level of patient-reported COC and patient satisfaction. It is necessary to adjust for unobserved confounders to obtain more accurate estimates of the relationship between patient-reported COC and patient satisfaction. However, the results and policy implications of this study should be cautiously interpreted because the possibility of other bias could not be ruled out. These findings support policies aimed at improving patient-reported COC among older adults.

Drug Enforcement Agency 2014 Hydrocodone Rescheduling Rule and Opioid Dispensing after Surgery

BACKGROUND

In 2014, the U.S. Drug Enforcement Agency reclassified hydrocodone from Schedule III to Schedule II of the Controlled Substances Act, resulting in new restrictions on refills. The authors hypothesized that hydrocodone rescheduling led to decreases in total opioid dispensing within 30 days of surgery and reduced new long-term opioid dispensing among surgical patients.

METHODS

The authors studied privately insured, opioid-naïve adults undergoing 10 general or orthopedic surgeries between 2011 and 2015. The authors conducted a differences-in-differences analysis that compared overall opioid dispensing before versus after the rescheduling rule for patients treated by surgeons who frequently prescribed hydrocodone before rescheduling (i.e., patients who were functionally exposed to rescheduling's impact) while adjusting for secular trends via a comparison group of patients treated by surgeons who rarely prescribed hydrocodone (i.e., unexposed patients). The primary outcome was any filled opioid prescription between 90 and 180 days after surgery; secondary outcomes included the 30-day refill rate and the amount of opioids dispensed initially and at 30 days postoperatively.

RESULTS

The sample included 65,136 patients. The percentage of patients filling a prescription beyond 90 days was similar after versus before rescheduling (absolute risk difference, -1.1%; 95% CI, -2.3% to 0.1%; P = 0.084). The authors estimated the rescheduling rule to be associated with a 45.4-mg oral morphine equivalent increase (difference-in-differences estimate; 95% CI, 34.2-56.7 mg; P < 0.001) in initial opioid dispensing, a 4.1% absolute decrease (95% CI, -5.5% to -2.7%; P < 0.001) in refills within 30 days, and a 37.7-mg oral morphine equivalent increase (95% CI, 20.6-54.8 mg; P = 0.008) in opioids dispensed within 30 days.

CONCLUSIONS

Among patients treated by surgeons who frequently prescribed hydrocodone before the Drug Enforcement Agency 2014 hydrocodone rescheduling rule, rescheduling did not impact long-term opioid receipt, although it was associated with an increase in opioid dispensing within 30 days of surgery.

Association of Physician Variation in Use of Manual Aspiration Thrombectomy With Outcomes Following Primary Percutaneous Coronary Intervention for ST-Elevation Myocardial Infarction: The National Cardiovascular Data Registry CathPCI Registry

Importance

Following negative randomized clinical trials, US guidelines downgraded support for routine manual aspiration thrombectomy (AT) during primary percutaneous coronary intervention (pPCI) for ST-segment elevation myocardial infarction (STEMI). However, some PCI operators continue to endorse a clinical benefit with AT use despite the lack of supportive data.

Objective

To examine temporal trends and comparative outcomes of AT use during pPCI for STEMI.

Design, Setting, and Participants

Retrospective cohort study of the National Cardiovascular Data Registry (NCDR) CathPCI Registry from July 1, 2009, to June 30, 2016, to assess temporal trends and in-hospital outcomes associated with AT use. To evaluate outcomes through 180 days, a subanalysis was conducted among Centers for Medicare and Medicaid Services-linked patients from July 1, 2009, through December 31, 2014. The comparative effectiveness analysis was performed using instrumental variable analyses to account for treatment selection bias. The instrumental variable was operator's preference to use AT during pPCI. Data were analyzed between February 1, 2017, and April 1, 2018.

Exposures

Aspiration thrombectomy use during pPCI for STEMI.

Main Outcomes and Measures

Primary outcomes included in-hospital stroke and death. Secondary outcomes included heart failure, stroke, all-cause rehospitalization, and death through 180 days of follow-up.

Results

Among all pPCIs performed (683 584), the mean (SD) age of patients was 61.7 (12.8) years, 489 257 were male (71.6%), and 596 384 were white (87.2%). Among patients undergoing pPCI, AT use increased from 2009 through 2011, with peak use of 13.8%. This was followed by a decline of more than 9%, reaching 4.7% by mid-2016. Overall, AT was used in 10.8% of pPCIs (lowest operator group median, 0%; highest operator group median, 33.8%). After instrumental variable analysis, AT use was associated with no difference in in-hospital death (adjusted absolute risk difference, -0.18%; 95% CI, -0.53% to 0.16%; P = .29) and a small increase in in-hospital stroke (adjusted RD, 0.14%; 95% CI, 0.01%-0.30%; P = .03). Among Centers for Medicare and Medicaid Services-linked patients, AT use was not associated with differences in death, heart failure, stroke, or rehospitalization at 180 days.

Conclusions and Relevance

In this large, nationwide analysis, AT use during STEMI pPCI declined by more than 50% since 2011, with use as of mid-2016 at less than 5%. Selective AT use was associated with a small excess risk of in-hospital stroke and no difference in other outcomes through 180 days of follow-up.

Instrumental Variable Analyses in Pharmacoepidemiology: What Target Trials Do We Emulate?

Purpose of review

When leveraging observational data to estimate treatment effects, it is useful to explicitly specify the “target trial” the investigators aspire to emulate. One concern is whether a proposed analysis plan can address the realities of the differences between the available non-randomized observational study and the target trial. When large or unknown sources of unmeasured confounding are suspected, investigators might consider turning to instrumental variable (IV) methods. Of course, the interpretation and appropriateness of IV analyses need to be considered carefully. The purpose of this review is to summarize recent methodologic advancements in how epidemiologists weigh the validity of an IV analysis and to place these methodologic advancements in the context of the feasible target trial’s protocol components.

Recent findings

There have been increased development and application of tools for sensitivity analyses, falsification strategies, and the identification of previously overlooked problems with IV analyses as applied in pharmacoepidemiology. Many of these recent insights can be seen as articulating restrictions on or tradeoffs between the types of target trials that can be validly emulated when using a classical IV analysis.

Summary

Putting classical IV methods in the context of target trials underscores the importance of recent methodologic developments and, more generally, when and how an IV analysis would be appropriate. We see that some tradeoffs in defining the target trials are unavoidable, that some tradeoffs may be offset or explored via sensitivity analyses, and that this serves as a framework for scientific discourse regarding IV and non-IV results emulating potentially different trials with different tradeoffs.

Evaluating different physician's prescribing preference based instrumental variables in two primary care databases: a study of inhaled long-acting beta2-agonist use and the risk of myocardial infarction

PURPOSE

Instrumental variable (IV) analysis with physician's prescribing preference (PPP) as IV is increasingly used in pharmacoepidemiology. However, it is unclear whether this IV performs consistently across databases. We aimed to evaluate the validity of different PPPs in a study of inhaled long-acting beta2-agonist (LABA) use and myocardial infarction (MI).

METHODS

Information on adults with asthma and/or COPD and at least one prescription of beta2-agonist, or muscarinic antagonist was extracted from the CPRD (UK) and the Mondriaan (Netherlands) databases. LABA exposure was considered time-fixed or time-varying. We measured PPPs using previous LABA prescriptions of physicians or proportion of LABA prescriptions per practice. Correlation (r) and standardized difference (SDif) were used to assess assumption of IV analysis.

RESULTS

For time-fixed LABA, the IV based on 10 previous prescriptions outperformed the other IVs regarding strength of the IV (r ≥ 0.15) and balance of confounders between IV categories (SDif < 0.10). None of the IVs we considered appeared to be valid for time-varying LABA. In CPRD (n = 490,499), which included approximately 18 times more subjects than Mondriaan (n = 27,459), IVs appeared more valid. LABA was not associated with MI; hazard ratios ranged from 0.86 to 1.18 for conventional analysis, and from 0.61 to 1.24 for the IV analyses with apparent valid IVs.

CONCLUSIONS

The validity of physician's prescribing preference as IV strongly depends on how this IV is defined and in which database it is applied. Hence, general recommendations cannot be made, other than to generate several plausible IVs, assess their validity, and report the estimate(s) from apparently valid IVs.

The missing cause approach to unmeasured confounding in pharmacoepidemiology

Unmeasured confounding is a major threat to the validity of pharmacoepidemiological studies of medication safety and effectiveness. We propose a new method for detecting and reducing the impact of unobserved confounding in large observational database studies. The method uses assumptions similar to the prescribing preference-based instrumental variable (IV) approach. Our method relies on the new 'missing cause' principle, according to which the impact of unmeasured confounding by (contra-)indication may be detected by assessing discrepancies between the following: (i) treatment actually received by individual patients and (ii) treatment that they would be expected to receive based on the observed data. Specifically, we use the treatment-by-discrepancy interaction to test for the presence of unmeasured confounding and correct the treatment effect estimate for the resulting bias. Under standard IV assumptions, we first proved that unmeasured confounding induces a spurious treatment-by-discrepancy interaction in risk difference models for binary outcomes and then simulated large pharmacoepidemiological studies with unmeasured confounding. In simulations, our estimates had four to six times smaller bias than conventional treatment effect estimates, adjusted only for measured confounders, and much smaller variance inflation than unbiased but very unstable IV estimates, resulting in uniformly lowest root mean square errors. The much lower variance of our estimates, relative to IV estimates, was also observed in an application comparing gastrointestinal safety of two classes of anti-inflammatory drugs. In conclusion, our missing cause-based method may complement other methods and enhance accuracy of analyses of large pharmacoepidemiological studies.

Potentially Inappropriate Medication and Health Care Outcomes: An Instrumental Variable Approach

OBJECTIVE

To examine the effects of potentially inappropriate medication (PIM) use on health care outcomes in elderly individuals using an instrumental variable (IV) approach.

DATA SOURCES/STUDY SETTING

Representative claim data from the universal health insurance program in Taiwan from 2007 to 2010.

STUDY DESIGN

We employed a panel study design to examine the relationship between PIM and hospitalization. We applied both the naive generalized estimating equation (GEE) model, which controlled for the observed patient and hospital characteristics, and the two-stage residual inclusion (2SRI) GEE model, which further accounted for the unobserved confounding factors. The PIM prescription rate of the physician most frequently visited by each patient was used as the IV.

PRINCIPAL FINDINGS

The naive GEE models indicated that patient PIM use was associated with a higher likelihood of hospitalization (odds ratio [OR], 1.399; 95 percent confidence interval [CI], 1.363-1.435). Using the physician PIM prescribing rate as an IV, we identified a stronger significant association between PIM and hospitalization (OR, 1.990; 95 percent CI, 1.647-2.403).

CONCLUSIONS

PIM use is associated with increased hospitalization in elderly individuals. Adjusting for unobserved confounders is needed to obtain unbiased estimates of the relationship between PIM and health care outcomes.

Terminal illness and the increased mortality risk of conventional antipsychotics in observational studies: a systematic review

INTRODUCTION

Numerous large observational studies have shown an increased risk of mortality in elderly users of conventional antipsychotics. Health authorities have warned against use of these drugs. However, terminal illness is a potentially strong confounder of the observational findings. So, the objective of this study was to systematically assess whether terminal illness may have biased the observational association between conventional antipsychotics and risk of mortality in elderly patients.

METHODS

Studies were searched in PubMed, CINAHL, Embase, the references of selected studies and articles referring to selected studies (Web of Science). Inclusion criteria were (i) observational studies that estimated (ii) the risk of all-cause mortality in (iii) new elderly users of (iv) conventional antipsychotics compared with atypical antipsychotics or no use. Two investigators assessed the characteristics of the exposure and reference groups, main results, measured confounders and methods used to adjust for unmeasured confounders.

RESULTS

We identified 21 studies. All studies were based on administrative medical and pharmaceutical databases. Sicker and older patients received conventional antipsychotics more often than new antipsychotics. The risk of dying was especially high in the first month of use, and when haloperidol was administered per injection or in high doses. Terminal illness was not measured in any study. Instrumental variables that were used were also confounded by terminal illness.

CONCLUSIONS

We conclude that terminal illness has not been adjusted for in observational studies that reported an increased risk of mortality risk in elderly users of conventional antipsychotics. As the validity of the evidence is questionable, so is the warning based on it.

Health Care Utilization and Treatment Persistence Associated with Oral Paliperidone and Lurasidone in Schizophrenia Treatment

BACKGROUND

Oral paliperidone and lurasidone are new second-generation antipsychotics (SGAs). Empirical evidence on the comparative costs and persistence of these 2 agents are absent in the literature.

OBJECTIVE

To assess health care use and persistence associated with the 2 new agents oral paliperidone and lurasidone and other SGAs.

METHODS

Schizophrenia patients who initiated SGA therapy were identified in the January 2007-June 2013 claims databases of a large managed care organization. Multivariate regressions using aripiprazole as the comparator were conducted. Ordinary least squares regressions were used to estimate the total medical and pharmacy costs associated with each drug. Poisson regressions were conducted to evaluate the frequency of hospitalizations and emergency department (ED) visits associated with each drug. A censored regression model was used to evaluate the comparative persistence. Sensitivity analyses using generalized linear models, two-part models, hurdle models, and instrumental variable regressions were also performed.  

RESULTS

Compared with aripiprazole, paliperidone was not associated with significantly different total costs, yet lurasidone was associated with lower total costs (-$7,052; 95% CI = -$9,221, -$4,882). Lurasidone was also associated with significantly lower medical services costs (-$5,025; 95% CI = -$7,096, -$2,955), drug costs (-$2,026; 95% CI = -$2,695, -$1,357), hospital costs (-$3,026; 95% CI = -$4,731, -$1,321), outpatient costs (-$1,999; 95% CI = -$2,536, -$1,463), and ED costs (-$2,284; 95% CI = -$3,069, -$1,499), whereas paliperidone did not have significant effects on any types of costs. Paliperidone users had fewer ED visits (-0.25; 95% CI = -0.42, -0.08), while lurasidone users had fewer hospitalizations (-5.98; 95% CI = -6.61, -5.35) and fewer ED visits (-2.51; 95% CI = -2.92, -2.10). Both paliperidone and lurasidone were associated with lower levels of treatment persistence.

CONCLUSIONS

Paliperidone does not associate with lower total costs compared with commonly used SGAs, whereas lurasidone is associated with lower total health costs. Thus, high access fees of lurasidone are not necessarily a major concern in prescription.

Risk of manic switch associated with antidepressant therapy in pediatric bipolar depression

OBJECTIVE

The purpose of this study was to assess the risk of manic switch associated with antidepressants in Medicaid-enrolled pediatric patients with bipolar depression.

METHODS

This retrospective cohort study involved 2003-2007 Medicaid Analytic eXtract (MAX) data from four geographically diverse states. The study sample included children and adolescents (ages 6-18 years) who had received a diagnosis of bipolar disorder on two or more separate occasions or during a hospital discharge, followed by a diagnosis of depression. According to the pharmacotherapy received by these patients in the 30 days around the index bipolar depression diagnosis, patients were categorized into five mutually exclusive groups. Manic switch was defined as having received a diagnosis of mania within 6 weeks after the initiation of bipolar depression treatment. Relative risks of manic switch between antidepressant monotherapy/polytherapy and their alternatives were assessed using Cox proportional hazards model. The robustness of the conventional Cox proportional hazards model toward possible bias caused by unobserved confounders was tested using instrumental variable analysis, and the uncertainty regarding manic switch definition was tested by altering the duration of follow-up.

RESULTS

After applying all the selection criteria, 179 antidepressant monotherapy, 1047 second-generation antipsychotic (SGA) monotherapy, 570 mood stabilizer monotherapy, 445 antidepressant polytherapy, and 1906 SGA-mood stabilizer polytherapy users were identified. In Cox proportional hazard analyses, both antidepressant monotherapy and polytherapy exhibited higher risk of manic switch than their alternatives (antidepressant monotherapy vs. SGA monotherapy, hazard ratio [HR]=2.87 [95% CI: 1.10-7.49]; antidepressant monotherapy vs. mood stabilizer monotherapy, HR=1.41 [95% CI: 0.52-3.80); antidepressant polytherapy vs. SGA-mood stabilizer polytherapy, HR=1.61 [95% CI: 0.90-2.89]). However, only the comparison between antidepressant monotherapy and SGA monotherapy was statistically significant. The instrumental variable analysis did not detect endogeneity of the treatment variables. Extending the follow-up period from 6 weeks to 8 and 12 weeks generated findings consistent with the main analysis.

CONCLUSIONS

Study findings indicated a higher risk of manic switch associated with antidepressant monotherapy than with SGA monotherapy in pediatric patients with bipolar depression. The finding supported the clinical practice of cautious prescribing of antidepressants for brief periods.

Capitalizing on prescribing pattern variation to compare medications for type 2 diabetes

BACKGROUND

Clinical trials often compare hypoglycemic medications on the basis of glycemic control but do not examine long-term outcomes (e.g., mortality). This study demonstrates an alternative approach to lengthening clinical trials to assess these long-term outcomes.

OBJECTIVE

To use observational quasi-experimental methods using instrumental variables (IVs) to compare the effect of two hypoglycemic medications, sulfonylureas (SUs) and thiazolidinediones (TZDs), on long-term outcomes.

METHODS

This study used administrative data from the Veterans Health Administration and Medicare from 2000 to 2010. The study population included US veterans dually enrolled in Medicare who received a prescription for metformin and then initiated SUs or TZDs. Patients could either continue on or discontinue metformin after the initiation of the second agent. Treatment was defined as starting either a SU or a TZD. Local variations in SU prescribing rates were used as instruments in IV models to control for selection bias. Survival models predicted all-cause mortality, ambulatory care sensitive condition hospitalizations, and stroke or heart attack (acute myocardial infarction).

RESULTS

Starting on SUs compared to TZDs significantly increased the likelihood of experiencing mortality and ACSC hospitalization. The estimated hazard ratio for the effect of starting on SUs compared to TZDs was 1.50 (95% confidence interval [CI] 1.09-2.09) for all-cause mortality, 1.68 (95% CI 1.31-2.15) for ambulatory care sensitive condition hospitalization, and 1.15 (95% CI 0.80-1.66) for acute myocardial infarction or stroke.

CONCLUSIONS

Our findings suggest increased risk of major adverse events associated with SUs as a second-line agent. Quasi-experimental IV methods may be an important alternative to lengthening clinical trials to assess long-term outcomes.

Adjusting for geographic variation in observational comparative effectiveness studies: a case study of antipsychotics using state Medicaid data

Background

Area-level variation in treatment and outcomes may be a potential source of confounding bias in observational comparative effectiveness studies. This paper demonstrates how to use exploratory spatial data analysis (ESDA) and spatial statistical methods to investigate and control for these potential biases. The case presented compares the effectiveness of two antipsychotic treatment strategies: oral second-generation antipsychotics (SGAs) vs. long-acting paliperiodone palmitate (PP).

Methods

A new-start cohort study was conducted analyzing patient-level administrative claims data (8/1/2008–4/30/2011) from Missouri Medicaid. ESDA techniques were used to examine spatial patterns of antipsychotic prescriptions and outcomes (hospitalization and emergency department (ED) visits). Likelihood of mental health-related outcomes were compared between patients starting PP (N = 295) and oral SGAs (N = 8,626) using multilevel logistic regression models adjusting for patient composition (demographic and clinical factors) and geographic region.

Results

ESDA indicated significant spatial variation in antipsychotic prescription patterns and moderate variation in hospitalization and ED visits thereby indicating possible confounding by geography. In the multilevel models for this antipsychotic case example, patient composition represented a stronger source of confounding than geographic context.

Conclusion

Because geographic variation in health care delivery is ubiquitous, it could be a comparative effectiveness research (CER) best practice to test for possible geographic confounding in observational data. Though the magnitude of the area-level geography effects were small in this case, they were still statistically significant and should therefore be examined as part of this observational CER study. More research is needed to better estimate the range of confounding due to geography across different types of observational comparative effectiveness studies and healthcare utilization outcomes.

Electronic supplementary material

The online version of this article (doi:10.1186/1472-6963-14-355) contains supplementary material, which is available to authorized users.

Practice-based versus patient-level outcomes research in hemodialysis: the DOPPS (Dialysis Outcomes and Practice Patterns Study) experience

When randomized controlled trials are unavailable, clinicians have to rely on observational studies. However, analyses using observational data to evaluate specific treatments and their associations with outcomes often are biased through confounding by clinical indication for the treatment of interest. Given the rich observational data and limited clinical trial data available in the dialysis population, successfully accounting for this bias can lead to substantial knowledge generation. In recent decades, much has been learned about statistical methods for observational data, including the fact that even extensive adjustments may not always overcome this bias, particularly when unmeasured confounders exist. In this article, examples based on the international DOPPS (Dialysis Outcomes and Practice Patterns Study) are used to demonstrate the value of practice-based instrumental variable analyses. This methodology leverages the marked differences in practice patterns among dialysis facilities and uses the reasonable assumption that patients are assigned to a dialysis facility without consideration of its specific treatment pattern in order to minimize bias in analyses relying on observational data. Examples using the dialysis facility as an instrument that are reviewed in depth in this article include studies of dialysate sodium concentration, systolic blood pressure targets, and treatment time, demonstrate the value of this methodology to produce advanced knowledge. However, practice-based analyses have potential limitations. Specifically, observation of sufficiently large differences in practice patterns is required and these analyses should consider that the treatment of interest may be associated with other facility treatment practices. These examples from the DOPPS hopefully will stimulate advances in methodologies and critical clinical work toward improving patient care by identifying beneficial treatment practices applicable to dialysis, chronic kidney disease, and beyond.

Comparative effectiveness of monotherapy with mood stabilizers versus second generation (atypical) antipsychotics for the treatment of bipolar disorder in children and adolescents

OBJECTIVE

This study compared the effectiveness and safety of second generation (atypical) antipsychotic (SGA) versus traditional mood stabilizers (MS) in children and adolescents with bipolar disorder.

METHODS

The study was a retrospective cohort study on 5 years (2003-2007) of Medicaid claims data from four geographically diversified states. Children and adolescents aged 6-18 years who initiated a new treatment episode for bipolar disorder on either an SGA or an MS were followed for 12 months to compare the effectiveness and safety between the two therapeutic categories for pediatric bipolar disorder (PBD). The outcome measures were psychiatric hospital admission, all cause medication discontinuation and treatment augmentation. Potential selection bias caused by unobserved confounding was addressed with instrumental variable methods, using physician prescribing preference and year of cohort entry as the instruments. Sensitivity analysis was conducted to test the robustness of findings against the uncertainties on PBD diagnosis.

RESULTS

Of the 7423 bipolar children and adolescents identified, 66.60% started treatment on SGA, whereas 33.40% initiated on MS. Patients who initiated on MS and SGA had comparable risk of psychiatric hospital admission (HR=1.172, 95%CI: 0.827-1.660). However, as compared with those who initiated on MS, patients who initiated on SGA were less likely to discontinue the treatment (HR=0.634, 95%CI: 0.419-0.961) and less likely to receive treatment augmentation (HR=0.223, 95%CI: 0.103-0.484).

CONCLUSION

As compared with MS monotherapy, SGA monotherapy could be a more effective and safer treatment option for PBD.

Comparing adherence to two different HIV antiretroviral regimens: an instrumental variable analysis

The objective of this observational cohort study was to compare adherence to protease inhibitor (PI)-based regimens or non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimens. HIV-seropositive, antiretroviral-naïve patients initiating therapy between 1998 and 2006 were identified using Veterans Health Administration databases. First-year adherence ratios were calculated as proportion of days covered (PDC). Multivariable regressions were run with an indicator for PDC >95, 90, 85, and 80 % as the dependent variable and an indicator for a PI-based regimen as the key independent variable. We controlled for residual unmeasured confounding by indication using an instrumental variable technique, using the physician's prescribing preference as the instrument. Out of 929 veterans on PI-based and 747 on NNRTI-based regimens, only 19.7 % of PI patients had PDC >80 %, compared to 35.1 % of NNRTI patients. In multivariable analysis, starting a PI regimen was significantly associated with poor adherence for all 4 adherence thresholds using conventional regressions and instrumental variable methods.

Longer dialysis session length is associated with better intermediate outcomes and survival among patients on in-center three times per week hemodialysis: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS)

BACKGROUND

Longer dialysis session length (treatment time, TT) has been associated with better survival among hemodialysis (HD) patients. The impact of TT on clinical markers that may contribute to this survival advantage is not well known.

METHODS

Using data from the international Dialysis Outcomes and Practice Patterns Study, we assessed the association of TT with clinical outcomes using both standard regression analyses and instrumental variable approaches. The study included 37,414 patients on in-center HD three times per week with prescribed TT from 120 to 420 min.

RESULTS

Facility mean TT ranged from 214 min in the USA to 256 min in Australia-New Zealand. Accounting for country effects, mortality risk was lower for patients with longer TT {hazard ratio for every 30 min: all-cause mortality: 0.94 [95% confidence interval (CI): 0.92-0.97], cardiovascular mortality: 0.95 (95% CI: 0.91-0.98) and sudden death: 0.93 (95% CI: 0.88-0.98)}. Patients with longer TT had lower pre- and post-dialysis systolic blood pressure, greater intradialytic weight loss, higher hemoglobin (for the same erythropoietin dose), serum albumin and potassium and lower serum phosphorus and white blood cell counts. Similar associations were found using the instrumental variable approach, although the positive associations of TT with weight loss and potassium were lost.

CONCLUSIONS

Favorable levels of a variety of clinical markers may contribute to the better survival of patients receiving longer TT. These findings support longer TT prescription in the setting of in-center, three times per week HD.

One-to-many propensity score matching in cohort studies

BACKGROUND

Among the large number of cohort studies that employ propensity score matching, most match patients 1:1. Increasing the matching ratio is thought to improve precision but may come with a trade-off with respect to bias.

OBJECTIVE

To evaluate several methods of propensity score matching in cohort studies through simulation and empirical analyses.

METHODS

We simulated cohorts of 20,000 patients with exposure prevalence of 10%-50%. We simulated five dichotomous and five continuous confounders. We estimated propensity scores and matched using digit-based greedy ("greedy"), pairwise nearest neighbor within a caliper ("nearest neighbor"), and a nearest neighbor approach that sought to balance the scores of the comparison patient above and below that of the treated patient ("balanced nearest neighbor"). We matched at both fixed and variable matching ratios and also evaluated sequential and parallel schemes for the order of formation of 1:n match groups. We then applied this same approach to two cohorts of patients drawn from administrative claims data.

RESULTS

Increasing the match ratio beyond 1:1 generally resulted in somewhat higher bias. It also resulted in lower variance with variable ratio matching but higher variance with fixed. The parallel approach generally resulted in higher mean squared error but lower bias than the sequential approach. Variable ratio, parallel, balanced nearest neighbor matching generally yielded the lowest bias and mean squared error.

CONCLUSIONS

1:n matching can be used to increase precision in cohort studies. We recommend a variable ratio, parallel, balanced 1:n, nearest neighbor approach that increases precision over 1:1 matching at a small cost in bias.

Blood pressure levels and mortality risk among hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study

KDOQI practice guidelines recommend predialysis blood pressure <140/90 mm Hg; however, most prior studies had found elevated mortality with low, not high, systolic blood pressure. This is possibly due to unmeasured confounders affecting systolic blood pressure and mortality. To lessen this bias, we analyzed 24,525 patients by Cox regression models adjusted for patient and facility characteristics. Compared with predialysis systolic blood pressure of 130-159 mm Hg, mortality was 13% higher in facilities with 20% more patients at systolic blood pressure of 110-129 mm Hg and 16% higher in facilities with 20% more patients at systolic blood pressure of ≥160 mm Hg. For patient-level systolic blood pressure, mortality was elevated at low (<130 mm Hg), not high (≥180 mm Hg), systolic blood pressure. For predialysis diastolic blood pressure, mortality was lowest at 60-99 mm Hg, a wide range implying less chance to improve outcomes. Higher mortality at systolic blood pressure of <130 mm Hg is consistent with prior studies and may be due to excessive blood pressure lowering during dialysis. The lowest risk facility systolic blood pressure of 130-159 mm Hg indicates this range may be optimal, but may have been influenced by unmeasured facility practices. While additional study is needed, our findings contrast with KDOQI blood pressure targets, and provide guidance on optimal blood pressure range in the absence of definitive clinical trial data.

Effect of colonoscopy on colorectal cancer incidence and mortality: an instrumental variable analysis

BACKGROUND

Using population-based health services information to estimate the effectiveness of colonoscopy on colorectal cancer (CRC) outcomes is prone to selection bias.

OBJECTIVE

To determine the effect of colonoscopy on CRC incidence and mortality.

DESIGN

Population-based retrospective cohort study.

SETTING

Ontario provincial health data information.

PATIENTS

This study involved average-risk persons aged 50 to 74 years from 1996 to 2000 who were alive and free of CRC on January 1, 2001.

INTERVENTION

Colonoscopy between 1996 and 2000.

MAIN OUTCOME MEASUREMENTS

CRC incidence and mortality from 2001 to 2007.

RESULTS

The study cohort contained 1,089,998 persons, 7.9% of whom had undergone a colonoscopy between 1996 and 2000. Using primary care physician rate of discretionary colonoscopy as an instrumental variable, the receipt of colonoscopy was associated with a 0.60% (95% confidence interval [CI], 0.31%-0.78%) absolute reduction in the 7-year colorectal cancer incidence and a 0.17% (95% CI, 0.14%-0.21%) absolute reduction in the 5-year risk of death caused by CRC. This corresponds to a 48% relative decrease in CRC incidence (risk ratio [RR] 0.52; 95% CI, 0.34-0.76) and 81% decrease in mortality caused by CRC (RR 0.19, 95% CI, 0.07-0.47). In subgroup analyses, the reduction in the risk of death due to CRC was larger in women than men. The reduction in CRC incidence was larger for complete colonoscopies and for left-sided cancers.

LIMITATIONS

Instrumental variable methods estimate only the marginal effect on the population studied.

CONCLUSION

Increased use of colonoscopy procedures is associated with a reduction in the incidence and mortality of CRC in the population studied.

Impact of physician preferences for homeopathic or conventional medicines on patients with musculoskeletal disorders: results from the EPI3-MSD cohort

OBJECTIVE

The objective of this study was to assess the effect of physician practicing preferences (PPP) in primary care for homeopathy (Ho), CAM (Complementary and alternative medicines) with conventional medicine (Mx) or exclusively conventional medicine (CM) on patients with musculoskeletal disorders (MSDs), with reference to clinical progression, drug consumption, side effects and loss of therapeutic opportunity.

METHODS

The EPI3-MSD study was a nationwide observational cohort of a representative sample of general practitioners (GP) and their patients in France. Recruitment of GP was stratified by PPP, which was self-declared. Diagnoses and comorbidities were recorded by GP at inclusion. Patients completed a standardized telephone interview at inclusion, one, three and twelve months, including MSD-functional scales and medication consumption.

RESULTS

1153 MSD patients were included in the three PPP groups. Patients did not differ between groups except for chronicity of MSDs (>12 weeks), which was higher in the Ho group (62.1%) than in the CM (48.6%) and Mx groups (50.3%). The twelve-month development of specific functional scores was identical across the three groups after controlling for baseline score (p > 0.05). After adjusting for propensity scores, NSAID use over 12 months was almost half in the Ho group (OR, 0.54; 95%CI, 0.38-0.78) as compared to the CM group; no difference was found in the Mx group (OR, 0.81; 95% CI: 0.59-1.15).

CONCLUSION

MSD patients seen by homeopathic physicians showed a similar clinical progression when less exposed to NSAID in comparison to patients seen in CM practice, with fewer NSAID-related adverse events and no loss of therapeutic opportunity.

Phosphate binder use and mortality among hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS): evaluation of possible confounding by nutritional status

BACKGROUND

Poor nutritional status and both hyper- and hypophosphatemia are associated with increased mortality in maintenance hemodialysis (HD) patients. We assessed associations of phosphate binder prescription with survival and indicators of nutritional status in maintenance HD patients.

STUDY DESIGN

Prospective cohort study (DOPPS [Dialysis Outcomes and Practice Patterns Study]), 1996-2008.

SETTING & PARTICIPANTS

23,898 maintenance HD patients at 923 facilities in 12 countries.

PREDICTORS

Patient-level phosphate binder prescription and case-mix-adjusted facility percentage of phosphate binder prescription using an instrumental-variable analysis.

OUTCOME

All-cause mortality.

RESULTS

Overall, 88% of patients were prescribed phosphate binders. Distributions of age, comorbid conditions, and other characteristics showed small differences between facilities with higher and lower percentages of phosphate binder prescription. Patient-level phosphate binder prescription was associated strongly at baseline with indicators of better nutrition, ie, higher values for serum creatinine, albumin, normalized protein catabolic rate, and body mass index and absence of cachectic appearance. Overall, patients prescribed phosphate binders had 25% lower mortality (HR, 0.75; 95% CI, 0.68-0.83) when adjusted for serum phosphorus level and other covariates; further adjustment for nutritional indicators attenuated this association (HR, 0.88; 95% CI, 0.80-0.97). However, this inverse association was observed for only patients with serum phosphorus levels ≥3.5 mg/dL. In the instrumental-variable analysis, case-mix-adjusted facility percentage of phosphate binder prescription (range, 23%-100%) was associated positively with better nutritional status and inversely with mortality (HR for 10% more phosphate binders, 0.93; 95% CI, 0.89-0.96). Further adjustment for nutritional indicators reduced this association to an HR of 0.95 (95% CI, 0.92-0.99).

LIMITATIONS

Results were based on phosphate binder prescription; phosphate binder and nutritional data were cross-sectional; dietary restriction was not assessed; observational design limits causal inference due to possible residual confounding.

CONCLUSIONS

Longer survival and better nutritional status were observed for maintenance HD patients prescribed phosphate binders and in facilities with a greater percentage of phosphate binder prescription. Understanding the mechanisms for explaining this effect and ruling out possible residual confounding require additional research.

Applying propensity scores estimated in a full cohort to adjust for confounding in subgroup analyses

BACKGROUND: A correctly specified propensity score (PS) estimated in a cohort ("cohort PS") should, in expectation, remain valid in a subgroup population. OBJECTIVE: We sought to determine whether using a cohort PS can be validly applied to subgroup analyses and, thus, add efficiency to studies with many subgroups or restricted data. METHODS: In each of three cohort studies, we estimated a cohort PS, defined five subgroups, and then estimated subgroup-specific PSs. We compared difference in treatment effect estimates for subgroup analyses adjusted by cohort PSs versus subgroup-specific PSs. Then, over 10 million times, we simulated a population with known characteristics of confounding, subgroup size, treatment interactions, and treatment effect and again assessed difference in point estimates. RESULTS: We observed that point estimates in most subgroups were substantially similar with the two methods of adjustment. In simulations, the effect estimates differed by a median of 3.4% (interquartile (IQ) range 1.3-10.0%). The IQ range exceeded 10% only in cases where the subgroup had < 1000 patients or few outcome events. CONCLUSIONS: Our empirical and simulation results indicated that using a cohort PS in subgroup analyses was a feasible approach, particularly in larger subgroups. Copyright © 2011 John Wiley & Sons, Ltd.

Opportunities for minimization of confounding in observational research

Observational epidemiological studies are increasingly used in pharmaceutical research to evaluate the safety and effectiveness of medicines. Such studies can complement findings from randomized clinical trials by involving larger and more generalizable patient populations by accruing greater durations of follow-up and by representing what happens more typically in the clinical setting. However, the interpretation of exposure effects in observational studies is almost always complicated by non-random exposure allocation, which can result in confounding and potentially lead to misleading conclusions. Confounding occurs when an extraneous factor, related to both the exposure and the outcome of interest, partly or entirely explains the relationship observed between the study exposure and the outcome. Although randomization can eliminate confounding by distributing all such extraneous factors equally across the levels of a given exposure, methods for dealing with confounding in observational studies include a careful choice of study design and the possible use of advanced analytical methods. The aim of this paper is to introduce some of the approaches that can be used to help minimize the impact of confounding in observational research to the reader working in the pharmaceutical industry.

Instrumental variable meta-analysis of individual patient data: application to adjust for treatment non-compliance

Background

Intention-to-treat (ITT) is the standard data analysis method which includes all patients regardless of receiving treatment. Although the aim of ITT analysis is to prevent bias due to prognostic dissimilarity, it is also a counter-intuitive type of analysis as it counts patients who did not receive treatment, and may lead to "bias toward the null." As treated (AT) method analyzes patients according to the treatment actually received rather than intended, but is affected by the selection bias. Both ITT and AT analyses can produce biased estimates of treatment effect, so instrumental variable (IV) analysis has been proposed as a technique to control for bias when using AT data. Our objective is to correct for bias in non-experimental data from previously published individual patient data meta-analysis by applying IV methods

Methods

Center prescribing preference was used as an IV to assess the effects of methotrexate (MTX) in preventing debilitating complications of chronic graft-versus-host-disease (cGVHD) in patients who received peripheral blood stem cell (PBSCT) or bone marrow transplant (BMT) in nine randomized controlled trials (1107 patients). IV methods are applied using 2-stage logistic, 2-stage probit and generalized method of moments models.

Results

ITT analysis showed a statistically significant detrimental effect with the use of day 11 MTX, resulting in cGVHD odds ratio (OR) of 1.34 (95% CI 1.02-1.76). AT results showed no difference in the odds of cGVHD with the use of MTX [OR 1.31 (95%CI 0.99-1.73)]. IV analysis further corrected the results toward no difference in the odds of cGVHD between PBSCT vs. BMT, allowing for a possibility of beneficial effects of MTX in preventing cGVHD in PBSCT recipients (OR 1.14; 95%CI 0.83-1.56).

Conclusion

All instrumental variable models produce similar results. IV estimates correct for bias and do not exclude the possibility that MTX may be beneficial, contradicting the ITT analysis.

Comparative effectiveness research: Policy context, methods development and research infrastructure

Comparative effectiveness research (CER) has received substantial attention as a potential approach for improving health outcomes while lowering costs of care, and for improving the relevance and quality of clinical and health services research. The Institute of Medicine defines CER as 'the conduct and synthesis of systematic research comparing different interventions and strategies to prevent, diagnose, treat, and monitor health conditions. The purpose of this research is to inform patients, providers, and decision-makers, responding to their expressed needs, about which interventions are most effective for which patients under specific circumstances.' Improving the methods and infrastructure for CER will require sustained attention to the following issues: (1) Meaningful involvement of patients, consumers, clinicians, payers, and policymakers in key phases of CER study design and implementation; (2) Development of methodological 'best practices' for the design of CER studies that reflect decision-maker needs and balance internal validity with relevance, feasibility and timeliness; and (3) Improvements in research infrastructure to enhance the validity and efficiency with which CER studies are implemented. The approach to addressing each of these issues should be informed by the understanding that the primary purpose of CER is to help health care decision makers make informed clinical and health policy decisions.

Use of instrumental variable in prescription drug research with observational data: a systematic review

OBJECTIVE

Instrumental variable (IV) analysis may offer a useful approach to the problem of unmeasured confounding in prescription drug research if the IV is: (1) strongly and unbiasedly associated to treatment assignment; and (2) uncorrelated with factors predicting the outcome (key assumptions).

STUDY DESIGN AND METHODS

We conducted a systematic review of the use of IV methods in prescription drug research to identify the major types of IVs and the evidence for meeting IV assumptions. We searched MEDLINE, OVID, PsychoInfo, EconLit, and economic databases from 1961 to 2009.

RESULTS

We identified 26 studies. Most (n=16) were published after 2007. We identified five types of IVs: regional variation (n=8), facility-prescribing patterns (n=5), physician preference (n=8), patient history/financial status (n=3), and calendar time (n=4). Evidence supporting the validity of IV was inconsistent. All studies addressed the first IV assumption; however, there was no standard for demonstrating that the IV sufficiently predicted treatment assignment. For the second assumption, 23 studies provided explicit argument that IV was uncorrelated with the outcome, and 16 supported argument with empirical evidence.

CONCLUSIONS

Use of IV methods is increasing in prescription drug research. However, we did not find evidence of a dominant IV. Future research should develop standards for reporting the validity and strength of IV according to key assumptions.

Comparative effectiveness of prostate cancer treatments: evaluating statistical adjustments for confounding in observational data

BACKGROUND

Using observational data to assess the relative effectiveness of alternative cancer treatments is limited by patient selection into treatment, which often biases interpretation of outcomes. We evaluated methods for addressing confounding in treatment and survival of patients with early-stage prostate cancer in observational data and compared findings with those from a benchmark randomized clinical trial.

METHODS

We selected 14 302 early-stage prostate cancer patients who were aged 66-74 years and had been treated with radical prostatectomy or conservative management from linked Surveillance, Epidemiology, and End Results-Medicare data from January 1, 1995, through December 31, 2003. Eligibility criteria were similar to those from a clinical trial used to benchmark our analyses. Survival was measured through December 31, 2007, by use of Cox proportional hazards models. We compared results from the benchmark trial with results from models with observational data by use of traditional multivariable survival analysis, propensity score adjustment, and instrumental variable analysis.

RESULTS

Prostate cancer patients receiving conservative management were more likely to be older, nonwhite, and single and to have more advanced disease than patients receiving radical prostatectomy. In a multivariable survival analysis, conservative management was associated with greater risk of prostate cancer-specific mortality (hazard ratio [HR] = 1.59, 95% confidence interval [CI] = 1.27 to 2.00) and all-cause mortality (HR = 1.47, 95% CI = 1.35 to 1.59) than radical prostatectomy. Propensity score adjustments resulted in similar patient characteristics across treatment groups, although survival results were similar to traditional multivariable survival analyses. Results for the same comparison from the instrumental variable approach, which theoretically equalizes both observed and unobserved patient characteristics across treatment groups, differed from the traditional multivariable and propensity score results but were consistent with findings from the subset of elderly patient with early-stage disease in the trial (ie, conservative management vs radical prostatectomy: for prostate cancer-specific mortality, HR = 0.73, 95% CI = 0.08 to 6.73; for all-cause mortality, HR = 1.09, 95% CI = 0.46 to 2.59).

CONCLUSION

Instrumental variable analysis may be a useful technique in comparative effectiveness studies of cancer treatments if an acceptable instrument can be identified.

Physical exercise among participants in the Dialysis Outcomes and Practice Patterns Study (DOPPS): correlates and associated outcomes

BACKGROUND

Levels of physical exercise among haemodialysis patients are low. Increased physical activity in this population has been associated with improved health-related quality of life (HRQoL) and survival. However, results of previous studies may not be applicable to the haemodialysis population as a whole. The present study provides the first description of international patterns of exercise frequency and its association with exercise programmes and clinical outcomes among participants in the Dialysis Outcomes and Practice Patterns Study (DOPPS).

METHODS

Data from a cross section of 20,920 DOPPS participants in 12 countries between 1996 and 2004 were analysed. Regular exercise was defined as exercise frequency equal to or more than once/week based on patient self-report. Linear mixed models and logistic regression assessed associations of exercise frequency with HRQoL and other psychosocial variables. Mortality risk was calculated in Cox proportional hazard models using patient-level (patient self-reported exercise frequency) and facility-level (the dialysis facility percentage of regular exercisers) predictors.

RESULTS

Regular exercise frequency varied widely across countries and across dialysis facilities within a country. Overall, 47.4% of participants were categorized as regular exercisers. The odds of regular exercise was 38% higher for patients from facilities offering exercise programmes (adjusted odds ratio = 1.38 [95% confidence interval: 1.03-1.84]; P = 0.03). Regular exercisers had higher HRQoL, physical functioning and sleep quality scores; reported fewer limitations in physical activities; and were less bothered by bodily pain or lack of appetite (P <or= 0.0001 for all). Regular exercise was also correlated with more positive patient affect and fewer depressive symptoms (P <or= 0.0001). In models extensively adjusted for demographics, comorbidities and socio-economic indicators, mortality risk was lower among regular exercisers (hazard ratio = 0.73 [0.69-0.78]; P < 0.0001) and at facilities with more regular exercisers (0.92 [0.89-0.94]; P < 0.0001 per 10% more regular exercisers).

CONCLUSIONS

Results from an international study of haemodialysis patients indicate that regular exercise is associated with better outcomes in this population and that patients at facilities offering exercise programmes have higher odds of exercising. Dialysis facility efforts to increase patient physical activity may be beneficial.

Instrumental variables II: instrumental variable application-in 25 variations, the physician prescribing preference generally was strong and reduced covariate imbalance

OBJECTIVE

An instrumental variable (IV) is an unconfounded proxy for a study exposure that can be used to estimate a causal effect in the presence of unmeasured confounding. To provide reliably consistent estimates of effect, IVs should be both valid and reasonably strong. Physician prescribing preference (PPP) is an IV that uses variation in doctors' prescribing to predict drug treatment. As reduction in covariate imbalance may suggest increased IV validity, we sought to examine the covariate balance and instrument strength in 25 formulations of the PPP IV in two cohort studies.

STUDY DESIGN AND SETTING

We applied the PPP IV to assess antipsychotic medication (APM) use and subsequent death among two cohorts of elderly patients. We varied the measurement of PPP, plus performed cohort restriction and stratification. We modeled risk differences with two-stage least square regression. First-stage partial r(2) values characterized the strength of the instrument. The Mahalanobis distance summarized balance across multiple covariates.

RESULTS

Partial r(2) ranged from 0.028 to 0.099. PPP generally alleviated imbalances in nonpsychiatry-related patient characteristics, and the overall imbalance was reduced by an average of 36% (+/-40%) over the two cohorts.

CONCLUSION

In our study setting, most of the 25 formulations of the PPP IV were strong IVs and resulted in a strong reduction of imbalance in many variations. The association between strength and imbalance was mixed.

Instrumental variables I: instrumental variables exploit natural variation in nonexperimental data to estimate causal relationships

The gold standard of study design for treatment evaluation is widely acknowledged to be the randomized controlled trial (RCT). Trials allow for the estimation of causal effect by randomly assigning participants either to an intervention or comparison group; through the assumption of "exchangeability" between groups, comparing the outcomes will yield an estimate of causal effect. In the many cases where RCTs are impractical or unethical, instrumental variable (IV) analysis offers a nonexperimental alternative based on many of the same principles. IV analysis relies on finding a naturally varying phenomenon, related to treatment but not to outcome except through the effect of treatment itself, and then using this phenomenon as a proxy for the confounded treatment variable. This article demonstrates how IV analysis arises from an analogous but potentially impossible RCT design, and outlines the assumptions necessary for valid estimation. It gives examples of instruments used in clinical epidemiology and concludes with an outline on estimation of effects.

Risiken von Antipsychotika im Alter, speziell bei Demenzen 1Prof. Dr. Hans Gutzmann zum 60. Geburtstag gewidmet

Antipsychotika werden häufig zur Behandlung von herausforderndem Verhalten bei Demenz eingesetzt. Antipsychotika der zweiten Generation («atypische Neuroleptika») spielen dabei mittlerweile eine größere Rolle als die älteren Substanzen. Der nur mäßigen Wirksamkeit stehen schwerwiegende Risiken gegenüber: so ist unter allen Antipsychotika die Mortalität insgesamt erhöht, aber auch das Risiko für plötzlichen Herztod, Schlaganfallereignisse oder venöse Thrombosen. Die Unterschiede zwischen Antipsychotika der ersten und der zweiten Generation sind dabei gering: hochpotente Antipsychotika der ersten Generation führen häufiger zu extrapyramidalmotorischen Symptomen, unter Antipsychotika der zweiten Generation ist das allgemeine Mortalitätsrisiko wahrscheinlich etwas geringer, andererseits wahrscheinlich venöse Thrombosen und metabolische Nebenwirkungen häufiger. Für die übrigen Nebenwirkungen bestehen keine gesicherten Unterschiede. Angesichts der großen Heterogenität der Antipsychotika erscheint der Nutzen einer pauschalen Aufteilung in Antipsychotika der ersten und zweiten Generation fragwürdig. Für die Auswahl einer Substanz sind substanzspezifische Nebenwirkungen, Dosierung, Aufdosierungsgeschwindigkeit sowie Beachtung von Komorbidität, Kontraindikationen und Arzneimittelinteraktionen in Relation zum individuellen Patienten erheblich wichtiger als die Wahl nach bloßer pauschaler Zuordnung zu einer dieser beiden Gruppen. Am wichtigsten ist aber, dass Antipsychotika nur bei klarer Indikation zur Anwendung kommen!

Bridging bench and practice: translational research for schizophrenia and other psychotic disorders

Translational research is urgently needed to turn basic scientific discoveries into widespread health gains and nowhere are these needs greater than in conditions such as schizophrenia and other psychotic disorders. In this article, we discuss one type of translational research--called T1--which is needed to take advantage of developments in the basic neurosciences and translate them into more efficacious diagnostic, preventive, and therapeutic interventions. However, ensuring that interventions from T1 research actually benefit patients will require a second form of translational research--called T2--to turn innovations into everyday clinical practice and health decision-making. Recent examples of T1 and T2 research in schizophrenia and other psychotic disorders as well as strategies for better linking T1 and T2 research agendas are covered.

Designed delays versus rigorous pragmatic trials: lower carat gold standards can produce relevant drug evaluations

BACKGROUND

Centralized administrative databases enable low-cost pragmatic randomized trials (PRTs) of drug effectiveness and safety. We simplified the PRT strategy by using designed delays (DD) to evaluate drug policies.

OBJECTIVES

To reassess our DD trial of a cost-saving nebulizer-to-inhaler conversion policy and a proposed DD trial of reduced restrictions on Cox-2 inhibitors.

RESEARCH DESIGN

We randomized 52 pairs of communities and clusters of physician practices to the policy either on time or after a 6-month delay. Our 2-stage qualitative reassessment comprised: (1) applying criteria for reporting PRTs and (2) assessing DD trials in 3 domains of responsibility: policymakers' decisions, researchers' decisions, and joint decisions involving negotiation.

MEASURES

A draft checklist of 22 Consolidated Standards of Reporting Trials (CONSORT). Researchers' recollections of their degree of influence on decisions.

RESULTS

DD trials deviated from ideal PRTs in the policymakers' domain: the policies affected mixtures of drugs, users, and illnesses, and implementation was not by strict protocol. Aspects negotiated by researchers and policymakers also deviated from ideal: length of delay; size and location of control group; unit of randomization; additional data collection; and communications to physicians. The DD trials complied better with CONSORT in the researchers' domain of analysis and interpretation.

CONCLUSIONS

DD trials can be negotiated with policymakers. Low cost and simplicity of DD trials partly compensate for some limitations for evaluating drug safety and effectiveness. The ethics question of whether a DD is routine evaluation or research depends on its purpose and generalizability.

Emerging Methods in Comparative Effectiveness and Safety

BACKGROUND

Interest in new methods for comparative effectiveness, drug and patient safety, and related studies is burgeoning. The advent of Medicare Part D for outpatient prescription drugs has drawn significant attention to the need for efficient ways to monitor the potential benefits and harms of pharmaceuticals. These trends prompted the Effective Health Care program at the Agency for Healthcare Research and Quality and its DEcIDE (Developing Evidence to Inform Decisions about Effectiveness) network to examine innovative approaches for such investigations through an invitational symposium in June 2006.

RESULTS

Conference papers covered numerous points about ways to structure both interventional and database-oriented studies, particularly those concerned with adverse drug events, to avoid bias in those studies, and to apply advanced statistical tools to exploit the information from these studies to their fullest. Of particular importance are: (1) using new types of experimental designs, including cluster randomization, delayed designs, pragmatic trials, and practice-based investigations that incorporate the natural variation of data from routine clinical practice; (2) finding efficient ways to use different types of databases-eg, Department of Veterans Affairs files, Centers for Disease Control and Prevention surveillance files, Medicaid claims data, and state hospital data-for examining initiation, persistence, and adherence, and the benefits and adverse events of pharmaceutical use; and (3) inventing or refining ways to decrease the threats to validity of analyses relying on administrative or other observational data, particularly through propensity scoring, inverse probability weighting, risk adjustment, and direct or indirect methods for synthesizing comparative effectiveness information.