Prior event rate ratio adjustment for hidden confounding in observational studies of treatment effectiveness: a pairwise Cox likelihood approach

Two assumptions of the prior event rate ratio approach for controlling confounding can be evaluated by self-controlled case series and dynamic random intercept modeling

OBJECTIVES

The prior event rate ratio (PERR) is a recently developed approach for controlling confounding by measured and unmeasured covariates in real-world evidence research and observational studies. Despite its rising popularity in studies of safety and effectiveness of biopharmaceutical products, there is no guidance on how to empirically evaluate its model assumptions. We propose two methods to evaluate two of the assumptions required by the PERR, specifically, the assumptions that occurrence of outcome events does not alter the likelihood of receiving treatment, and that earlier event rate does not affect later event rate.

STUDY DESIGN AND SETTING

We propose using self-controlled case series (SCCS) and dynamic random intercept modeling (DRIM), respectively, to evaluate the two aforementioned assumptions. A nonmathematical introduction of the methods and their application to evaluate the assumptions are provided. We illustrate the evaluation with secondary analysis of deidentified data on pneumococcal vaccination and clinical pneumonia in The Gambia, West Africa.

RESULTS

SCCS analysis of data on 12,901 vaccinated Gambian infants did not reject the assumption of clinical pneumonia episodes had no influence on the likelihood of pneumococcal vaccination. DRIM analysis of 14,325 infants with a total of 1719 episodes of clinical pneumonia did not reject the assumption of earlier episodes of clinical pneumonia had no influence on later incidence of the disease.

CONCLUSION

The SCCS and DRIM methods can facilitate appropriate use of the PERR approach to control confounding.

PLAIN LANGUAGE SUMMARY

The prior event rate ratio is a promising approach for analysis of real-world data and observational studies. We propose two statistical methods to evaluate the validity of two assumptions it is based on. They can facilitate appropriate use of the prior even rate ratio.

Association of COVID-19 vaccination with risks of hospitalization due to cardiovascular and other diseases: A study using data from the UK Biobank

OBJECTIVES

To explore whether COVID-19 vaccination protects against hospital admission by preventing infections and severe disease.

METHODS

We leveraged the UK Biobank and studied associations of COVID-19 vaccination (BioNTech-BNT162b2 or Oxford-AstraZeneca-ChAdOx1) with hospitalizations from cardiovascular and other selected diseases (N = 393,544; median follow-up = 54 days among vaccinated individuals). Multivariable Cox, Poisson regression, propensity score matching, and inverse probability treatment weighting analyses were performed. We also performed adjustment using prescription-time distribution matching, and prior event rate ratio.

RESULTS

We observed that COVID-19 vaccination (at least one dose), compared with no vaccination, was associated with reduced short-term risks of hospitalizations from stroke (hazard ratio [HR] = 0.178, 95% confidence interval [CI]: 0.127-0.250, P = 1.50e-23), venous thromboembolism (HR = 0.426, CI: 0.270-0.673, P = 2.51e-4), dementia (HR = 0.114, CI: 0.060-0.216; P = 2.24e-11), non-COVID-19 pneumonia (HR = 0.108, CI: 0.080-0.145; P = 2.20e-49), coronary artery disease (HR = 0.563, CI: 0.416-0.762; P = 2.05e-4), chronic obstructive pulmonary disease (HR = 0.212, CI: 0.126-0.357; P = 4.92e-9), type 2 diabetes (HR = 0.216, CI: 0.096-0.486, P = 2.12e-4), heart failure (HR = 0.174, CI: 0.118-0.256, P = 1.34e-18), and renal failure (HR = 0.415, CI: 0.255-0.677, P = 4.19e-4), based on standard Cox regression models. Among the previously mentioned results, reduced hospitalizations for stroke, heart failure, non-COVID-19 pneumonia, and dementia were consistently observed across regression, propensity score matching/inverse probability treatment weighting, prescription-time distribution matching, and prior event rate ratio. The results for two-dose vaccination were similar.

CONCLUSIONS

Taken together, this study provides further support to the safety and benefits of COVID-19 vaccination, and such benefits may extend beyond reduction of infection risk or severity per se. However, causal relationship cannot be concluded and further studies are required.

Engagement with a youth violence intervention programme is associated with lower re-attendance after violent injury: A UK major trauma network observational study

The hospital based Redthread Youth Violence Intervention Programme (YVIP) utilises experienced youth workers to support 11-24 year olds following an episode of violent injury, assault or exploitation who present to the Emergency Department (ED) at the East Midlands Major Trauma Centre (MTC), Nottingham, UK. The YVIP aims to promote personal change with the aim of reducing the incidence of further similar events. We conducted a retrospective, observational, cohort study to examine the association between engagement with the YVIP and re-attendance rates to the ED following a referral to Redthread. We also examined factors associated with engagement with the full YVIP. We found that 573 eligible individuals were referred to the YVIP over two years. Assault with body parts 34.9% (n = 200) or a bladed object 29.8% (n = 171) were the commonest reason for referral. A prior event rate ratio (PERR) analysis was used to compare rates of attendance between those who did and did not engage with the full YVIP. Engagement was associated with a reduction in re-attendances of 51% compared to those who did not engage (PERR 0.49 [95% 0.28-0.64]). A previous attendance to the ED by an individual positively predicted engagement. (OR 2.82 [95% CI 1.07-7.42], P = 0.035). A weekend attendance (OR 0.26 [0.15-0.44], P<0.001) and a phone call approach (OR 0.25 [0.14-0.47], P = 0.001), rather than a face-to-face approach by a Redthread worker, negatively impacted engagement. In conclusion, assaults with or without a weapon contributed to a significant proportion of attendances among this age group. The Redthread YVIP was associated with reduced rates of re-attendance to the East Midlands MTC among young persons who engaged with the full programme.

Use of Bisphosphonates and the Risk of Skin Ulcer: A National Cohort Study Using Data from the French Health Care Claims Database

INTRODUCTION

Previous pre-clinical and pharmacovigilance disproportionality analyses highlighted a safety signal of cutaneous ulcer with bisphosphonate use. Therefore, our objective is to evaluate this risk and assess whether unmeasured confounding factors could explain this association.

METHODS

This study is a population-based cohort study from a representative sample (1/97th) of the French health insurance claims database: Echantillon Généraliste des Bénéficiaires (EGB) from 2006 to 2019. To limit the impact of our study design and methodological choices on any association between skin ulceration and exposure to bisphosphonates, we used several methods: a Cox proportional hazards analysis and a prior event rate ratio (PERR) analysis, using two propensity matched control groups, and either the first episode of incident ulceration or multiple event-time outcomes.

RESULTS

There were 7402 individuals newly exposed to bisphosphonates matched to 29,605 unexposed individuals on propensity score. The primary outcome was skin ulcer occurrence assessed by at least 2 deliveries of wound dressing during the period of one month. Among 6911 individuals newly exposed to bisphosphonates and 28,072 unexposed individuals with no previous skin ulcer, the Cox regression yielded a hazard ratio (HR) of 1.40 (95% CI 1.26-1.56) for newly exposed individuals. Among 7402 exposed and 29,605 unexposed individuals, the PERR analysis found a non-significant HR of 1.03 (95% CI 0.87-1.24). Results were similar on the different sensitivity analyses.

CONCLUSION

No association between bisphosphonate and skin ulcers was found in the French population. The association observed in previous pharmacovigilance studies and in the Cox regression analysis is likely due to unmeasured confounding factors.

Dealing with confounding in observational studies: A scoping review of methods evaluated in simulation studies with single-point exposure

The aim of this article was to perform a scoping review of methods available for dealing with confounding when analyzing the effect of health care treatments with single-point exposure in observational data. We aim to provide an overview of methods and their performance assessed by simulation studies indexed in PubMed. We searched PubMed for simulation studies published until January 2021. Our search was restricted to studies evaluating binary treatments and binary and/or continuous outcomes. Information was extracted on the methods' assumptions, performance, and technical properties. Of 28,548 identified references, 127 studies were eligible for inclusion. Of them, 84 assessed 14 different methods (ie, groups of estimators that share assumptions and implementation) for dealing with measured confounding, and 43 assessed 10 different methods for dealing with unmeasured confounding. Results suggest that there are large differences in performance between methods and that the performance of a specific method is highly dependent on the estimator. Furthermore, the methods' assumptions regarding the specific data features also substantially influence the methods' performance. Finally, the methods result in different estimands (ie, target of inference), which can even vary within methods. In conclusion, when choosing a method to adjust for measured or unmeasured confounding it is important to choose the most appropriate estimand, while considering the population of interest, data structure, and whether the plausibility of the methods' required assumptions hold.

Influenza vaccination reduced myocardial infarctions in United Kingdom older adults: a prior event rate ratio study

OBJECTIVES

We aimed to estimate the real-world effectiveness of the influenza vaccine against myocardial infarction (MI) and influenza in the decade since adults aged ≥ 65 years were first recommended the vaccine.

STUDY DESIGN AND SETTING

We identified annual cohorts, 1997 to 2011, of adults aged ≥ 65 years, without previous influenza vaccination, from UK general practices, registered with the Clinical Practice Research Datalink. Using a quasi-experimental study design to control for confounding bias, we estimated influenza vaccine effectiveness on hospitalization for MI, influenza, and antibiotic prescriptions for lower respiratory tract infections.

RESULTS

Vaccination was moderately effective against influenza, the prior event rate ratio-adjusted hazard ratios ranging from 0.70 in 1999 to 0.99 in 2001. Prior event rate ratio-adjusted hazard ratios demonstrated a protective effect against MIs, varying between 0.40 in 2010 and 0.89 in 2001. Aggregated across the cohorts, influenza vaccination reduced the risk of MIs by 39% (95% confidence interval: 34%, 44%).

CONCLUSION

Effectiveness of the flu vaccine in preventing MIs in older UK adults is consistent with the limited evidence from clinical trials. Similar trends in effectiveness against influenza and against MIs suggest the risk of influenza mediates the effectiveness against MIs, although divergence in some years implies the mechanism may be complex.

Real-world effectiveness of pneumococcal vaccination in older adults: Cohort study using the UK Clinical Practice Research Datalink

BACKGROUND

The 23-valent pneumococcal polysaccharide vaccine (PPV23) is recommended for UK older adults, but how age moderates effectiveness is unclear.

METHODS

Three annual cohorts of primary-care patients aged≥65y from the Clinical Practice Research Datalink selected from 2003-5 created a natural experiment (n = 324,804), reflecting the staged introduction of the vaccine. The outcome was symptoms consistent with community-acquired pneumococcal pneumonia (CAP) requiring antibiotics or hospitalisation. We used the prior event rate ratio (PERR) approach to address bias from unmeasured confounders.

RESULTS

Vaccinated patients had higher rates of CAP in the year before vaccination than their controls, indicating the potential for confounding bias. After adjustment for confounding using the prior event rate ratio (PERR) method, PPV23 was estimated to be effective against CAP for two years after vaccination in all age sub-groups with hazard ratios (95% confidence intervals) of 0.86 (0.80 to 0.93), 0.74 (0.65 to 0.85) and 0.65 (0.57 to 0.74) in patients aged 65-74, 75-79 and 80+ respectively in the 2005 cohort. Age moderated the effect of vaccination with predicted risk reductions of 8% at 65y and 29% at 80y.

CONCLUSIONS

PPV23 is moderately effective at reducing CAP among UK patients aged≥65y, in the two years after vaccination. Vaccine effectiveness is maintained, and may increase, in the oldest age groups in step with increasing susceptibility to CAP.

The Effect of Buprenorphine on Human Immunodeficiency Virus Viral Suppression

BACKGROUND

Opioid use is prevalent among people living with human immunodeficiency virus (HIV; PLWH) and adversely affects HIV outcomes. We assessed the effect of buprenorphine (BUP) initiation on subsequent HIV viral loads.

METHODS

We identified PLWH from the Johns Hopkins HIV Clinical Cohort who initiated BUP between 2002 and 2017. Poisson regression with robust variance was used to estimate the prevalence of viral suppression (<200 copies/mL) before and after BUP initiation. We matched individuals who initiated BUP with controls based on viral load measurement dates and used prior event rate ratio (PERR) methods to estimate the effect of BUP initiation on viral suppression. PERR methods account for unmeasured confounders.

RESULTS

We identified 279 PLWH who initiated BUP. After BUP initiation, PLWH were more likely to be virally suppressed (prevalence ratio [PR], 1.19; 95% confidence interval [CI], 1.03-1.37). After matching PLWH who initiated BUP to controls and accounting for measured and unmeasured confounders, BUP initiation increased viral suppression for both those on antiretroviral therapy (ART) at baseline (PERR PR, 1.08; 95% CI, 1.00-1.18) and those not on ART at baseline (PR, 1.31; 95% CI, 1.10-1.61).

CONCLUSIONS

Our results indicate that the initiation of BUP results in an increase in the probability of being virally suppressed after accounting for both measured and unmeasured confounders. Persons with opioid use disorder should initiate BUP to not only treat substance use but also to increase viral suppression allowing for treatment as prevention.

Economic Assessment of High-Dose Versus Adjuvanted Influenza Vaccine: An Evaluation of Hospitalization Costs Based on a Cohort Study

Two influenza vaccines are licensed in the U.S. exclusively for the 65 years and older population: a trivalent inactivated high-dose influenza vaccine (HD-IIV3) and a trivalent inactivated adjuvanted influenza vaccine (aIIV3). In a recent publication, we estimated a relative vaccine effectiveness (rVE) of HD-IIV3 vs. aIIV3 of 12% (95% CI: 3.3–20%) for influenza-related hospitalizations using a retrospective study design, but did not report the number of prevented hospitalizations nor the associated avoided cost. In this paper we report estimations for both. Methods: Leveraging the rVE of a cohort study over two influenza seasons (2016/17 and 2017/18), we collected cost data for healthcare provided to the same study population. Vaccine costs were obtained from the Medicare pricing schedule. Our economic assessment compared cost of vaccination and hospital care for patients experiencing acute respiratory or cardiovascular illness. Results: We analyzed 1.9 million HD-IIV3 and 223,793 aIIV3 recipients. Average vaccine list prices were $46.23 for HD-IIV3 and $48.26 for aIIV3. The hospitalization rates for respiratory disease in HD-IIV3 and aIIV3 recipients were 187 (95% CI: 185–189) and 212 (195–231) per 10,000 persons-years, respectively. Attributing the average cost per hospitalization of $12,652 ($12,214–$13,090) to the difference in hospitalization rates, we estimate net savings of HD-IIV3 to be $34 ($10–$62) per recipient. Conclusion: Pooled over two predominantly A/H3N2 respiratory seasons, vaccination with HD-IIV3 was associated with lower hospitalization rates and associated costs compared to aIIV3 in senior members of a large national managed health care company in the U.S. Reduced hospitalizations affect healthcare utilization overall, and therefore other costly health outcomes.

Impact of influenza vaccination on amoxicillin prescriptions in older adults: A retrospective cohort study using primary care data

Background

Bacterial infections of the upper and lower respiratory tract are a frequent complication of influenza and contribute to the widespread use of antibiotics. Influenza vaccination may help reduce both appropriate and inappropriate prescribing of antibiotics. Electronic health records provide a rich source of information for assessing secondary effects of influenza vaccination.

Methods

We conducted a retrospective study to estimate effects of influenza vaccine on antibiotic (amoxicillin) prescription in the elderly based on data from the Clinical Practice Research Datalink. The introduction of UK policy to recommend the influenza vaccine to older adults in 2000 led to a substantial increase in uptake, creating a natural experiment. Of 259,753 eligible patients that were unvaccinated in 1999 and aged≥65y by January 2000, 88,519 patients received influenza vaccination in 2000. These were propensity score matched 1:1 to unvaccinated patients. Time-to-amoxicillin was analysed using the Prior Event Rate Ratio (PERR) Pairwise method to address bias from time-invariant measured and unmeasured confounders. A simulation study and negative control outcome were used to help strengthen the validity of results.

Results

Compared to unvaccinated patients, those from the vaccinated group were more likely to be prescribed amoxicillin in the year prior to vaccination: hazard ratio (HR) 1.90 (95% confidence interval 1.83, 1.98). Following vaccination, the vaccinated group were again more likely to be prescribed amoxicillin, HR 1.64 (1.58,1.71). After adjusting for prior differences between the two groups using PERR Pairwise, overall vaccine effectiveness was 0.86 (0.81, 0.92). Additional analyses suggested that provided data meet the PERR assumptions, these estimates were robust.

Conclusions

Once differences between groups were taken into account, influenza vaccine had a beneficial effect, lowering the frequency of amoxicillin prescribing in the vaccinated group. Ensuring successful implementation of national programmes of vaccinating older adults against influenza may help contribute to reducing antibiotic resistance.

On the Causal Interpretation of Rate-Change Methods: The Prior Event Rate Ratio and Rate Difference

A growing number of studies use data before and after treatment initiation in groups exposed to different treatment strategies to estimate “causal effects” using a ratio measure called the prior event rate ratio (PERR). Here, we offer a causal interpretation for PERR and its additive scale analog, the prior event rate difference (PERD). We show that causal interpretation of these measures requires untestable rate-change assumptions about the relationship between 1) the change of the counterfactual rate before and after treatment initiation in the treated group under hypothetical intervention to implement the control strategy; and 2) the change of the factual rate before and after treatment initiation in the control group. The rate-change assumption is on the multiplicative scale for PERR but on the additive scale for PERD; the 2 assumptions hold simultaneously under testable, but unlikely, conditions. Even if investigators can pick the most appropriate scale, the relevant rate-change assumption might not hold exactly, so we describe sensitivity analysis methods to examine how assumption violations of different magnitudes would affect study results. We illustrate the methods using data from a published study of proton pump inhibitors and pneumonia.

Precision Medicine in Type 2 Diabetes: Using Individualized Prediction Models to Optimize Selection of Treatment

Despite the known heterogeneity of type 2 diabetes and variable response to glucose lowering medications, current evidence on optimal treatment is predominantly based on average effects in clinical trials rather than individual-level characteristics. A precision medicine approach based on treatment response would aim to improve on this by identifying predictors of differential drug response for people based on their characteristics and then using this information to select optimal treatment. Recent research has demonstrated robust and clinically relevant differential drug response with all noninsulin treatments after metformin (sulfonylureas, thiazolidinediones, dipeptidyl peptidase 4 [DPP-4] inhibitors, glucagon-like peptide 1 [GLP-1] receptor agonists, and sodium-glucose cotransporter 2 [SGLT2] inhibitors) using routinely available clinical features. This Perspective reviews this current evidence and discusses how differences in drug response could inform selection of optimal type 2 diabetes treatment in the near future. It presents a novel framework for developing and testing precision medicine-based strategies to optimize treatment, harnessing existing routine clinical and trial data sources. This framework was recently applied to demonstrate that "subtype" approaches, in which people are classified into subgroups based on features reflecting underlying pathophysiology, are likely to have less clinical utility compared with approaches that combine the same features as continuous measures in probabilistic "individualized prediction" models.

Aspirin reduces cardiovascular events in patients with pneumonia: a prior event rate ratio analysis in a large primary care database

BACKGROUND

Ischaemic stroke and myocardial infarction (MI) are common after pneumonia and are associated with long-term mortality. Aspirin may attenuate this risk and should be explored as a therapeutic option.

METHODS

We extracted all patients with pneumonia (aged over 50 years) from the Clinical Practice Research Datalink (CPRD), a large UK primary care database, from inception until January 2019. We then performed a prior event rate ratio (PERR) analysis with propensity score matching (PSM), an approach that allows for control of measured and unmeasured confounding, with aspirin usage as the exposure and ischaemic events as the outcome. The primary outcome was the combined outcome of ischaemic stroke and MI. Secondary outcomes were ischaemic stroke and MI individually. Relevant confounders (smoking, comorbidities, age and gender) were included in the analysis.

FINDINGS

48 743 patients were eligible for matching. Of these, 9864 were aspirin users who were matched to 9864 non-users. Aspirin users had a reduced risk of the primary outcome (adjusted hazard ratio 0.64, 95% CI 0.52-0.79) in the PERR analysis. For both secondary outcomes, aspirin use was also associated with a reduced risk for MI (hazard ratio 0.46, 95% CI 0.30-0.72) and stroke (hazard ratio 0.70, 95% CI 0.55-0.91), respectively.

INTERPRETATION

This study provides supporting evidence that aspirin use is associated with reduced ischaemic events after pneumonia in a primary care setting. This drug may have a future clinical role in preventing this important complication.

Prior event rate ratio adjustment produced estimates consistent with randomized trial: a diabetes case study

OBJECTIVES

Electronic health records (EHR) provide a valuable resource for assessing drug side-effects, but treatments are not randomly allocated in routine care creating the potential for bias. We conduct a case study using the Prior Event Rate Ratio (PERR) Pairwise method to reduce unmeasured confounding bias in side-effect estimates for two second-line therapies for type 2 diabetes, thiazolidinediones, and sulfonylureas.

STUDY DESIGN AND SETTINGS

Primary care data were extracted from the Clinical Practice Research Datalink (n = 41,871). We utilized outcomes from the period when patients took first-line metformin to adjust for unmeasured confounding. Estimates for known side-effects and a negative control outcome were compared with the A Diabetes Outcome Progression Trial (ADOPT) trial (n = 2,545).

RESULTS

When on metformin, patients later prescribed thiazolidinediones had greater risks of edema, HR 95% CI 1.38 (1.13, 1.68) and gastrointestinal side-effects (GI) 1.47 (1.28, 1.68), suggesting the presence of unmeasured confounding. Conventional Cox regression overestimated the risk of edema on thiazolidinediones and identified a false association with GI. The PERR Pairwise estimates were consistent with ADOPT: 1.43 (1.10, 1.83) vs. 1.39 (1.04, 1.86), respectively, for edema, and 0.91 (0.79, 1.05) vs. 0.94 (0.80, 1.10) for GI.

CONCLUSION

The PERR Pairwise approach offers potential for enhancing postmarketing surveillance of side-effects from EHRs but requires careful consideration of assumptions.

Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network

Background: Unmeasured confounders are commonplace in observational studies conducted using real‐world data. Prior event rate ratio (PERR) adjustment is a technique shown to perform well in addressing such confounding. However, it has been demonstrated that, in some circumstances, the PERR method actually increases rather than decreases bias. In this work, we seek to better understand the robustness of PERR adjustment.

Methods: We begin with a Bayesian network representation of a generalized observational study, which is subject to unmeasured confounding. Previous work evaluating PERR performance used Monte Carlo simulation to calculate joint probabilities of interest within the study population. Here, we instead use a Bayesian networks framework.

Results: Using this streamlined analytic approach, we are able to conduct probabilistic bias analysis (PBA) using large numbers of combinations of parameters and thus obtain a comprehensive picture of PERR performance. We apply our methodology to a recent study that used the PERR in evaluating elderly‐specific high‐dose (HD) influenza vaccine in the US Veterans Affairs population. That study obtained an HD relative effectiveness of 25% (95% CI: 2%‐43%) against influenza‐ and pneumonia‐associated hospitalization, relative to standard‐dose influenza vaccine. In this instance, we find that the PERR‐adjusted result is more like to underestimate rather than to overestimate the relative effectiveness of the intervention.

Conclusions: Although the PERR is a powerful tool for mitigating the effects of unmeasured confounders, it is not infallible. Here, we develop some general guidance for when a PERR approach is appropriate and when PBA is a safer option.

Comparative effectiveness of high dose versus adjuvanted influenza vaccine: A retrospective cohort study

BACKGROUND

Adults 65 years and older (seniors) experience more complications following influenza infection than younger adults. We estimated the relative vaccine effectiveness (rVE) of a trivalent high dose (HD-IIV3) versus an adjuvanted trivalent influenza vaccine (aIIV3) in seniors for respiratory-related hospitalizations.

METHODS

We conducted a retrospective cohort study using claims data from Optum's Clinformatics® Data Mart to compare outcome rates between seniors who received HD-IIV3 versus aIIV3 during the 2016/17 and 2017/18, predominantly A/H3N2 respiratory seasons. Rates were adjusted for demographic characteristics, comorbid conditions, previous influenza vaccination, and geography. We used the previous event rate ratio (PERR) approach to address bias by time-fixed unmeasured confounders.

RESULTS

We identified 842,282 HD-IIV3 and 34,157 aIIV3 recipients for the 2016/17 season and 1,058,638 HD-IIV3 and 189,636 aIIV3 recipients for the 2017/18 season. The pooled rVE of HD-IIV3 versus aIIV3 for respiratory-related hospitalizations over both seasons was 12% (95% confidence interval: 3.3%-20%); 13% (-6.4% to 32%) for the 2016/17 season and 12% (2.1%-21%) for the 2017/18 season.

CONCLUSIONS

Pooled over two predominantly A/H3N2 respiratory seasons, HD-IIV3 was associated with fewer respiratory hospital admissions than aIIV3 in senior members of large national managed health care company in the U.S.

Relative Vaccine Effectiveness of High-Dose Versus Standard-Dose Influenza Vaccines Among Veterans Health Administration Patients

Background

We examined whether a high-dose inactivated influenza vaccine was more efficacious in preventing hospitalizations than a standard-dose vaccine in the Veterans Health Administration (VHA) senior population.

Methods

This study estimated the relative vaccine effectiveness (rVE) of high dose versus standard dose using a retrospective cohort of VHA patients 65 years of age or older in the 2015-2016 influenza season. To adjust for measured confounders, we matched each high-dose recipient with up to 4 standard-dose recipients vaccinated at the same location within a 2-week period and having 2 or more pre-existing medical comorbidities. We used the previous event rate ratio method (PERR), a type of difference-in-differences analysis, to adjust for unmeasured confounders.

Results

We evaluated 104965 standard-dose and 125776 high-dose recipients; matching decreased the population to 49091 standard-dose and 24682 high-dose recipients. The matched, PERR-adjusted rVE was 25% (95% confidence interval [CI], 2%-43%) against influenza- or pneumonia-associated hospitalization, 7% (95% CI, -2% to 14%) against all-cause hospitalization, 14% (95% CI, -8% to 32%) against influenza- or pneumonia-associated outpatient visit, 5% (95% CI, 2%-8%) against all-cause outpatient visit, and 38% (95% CI, -5% to 65%) against laboratory-confirmed influenza.

Conclusions

In protecting senior VHA patients against influenza- or pneumonia-associated hospitalization, a high-dose influenza vaccine is more effective than a standard-dose vaccine.

Proton-Pump Inhibitors and Long-Term Risk of Community-Acquired Pneumonia in Older Adults

OBJECTIVES

To estimate associations between long-term use of proton pump inhibitors (PPIs) and pneumonia incidence in older adults in primary care.

DESIGN

Longitudinal analyses of electronic medical records.

SETTING

England PARTICIPANTS: Individuals aged 60 and older in primary care receiving PPIs for 1 year or longer (N=75,050) and age- and sex-matched controls (N=75,050).

MEASUREMENTS

Net hazard ratios for pneumonia incidence in Year 2 of treatment were estimated using the prior event rate ratio (PERR), which adjusts for pneumonia incidence differences before initiation of treatment. Inverse probability weighted models adjusted for 78 demographic, disease, medication, and healthcare usage measures.

RESULTS

During the second year after initiating treatment, PPIs were associated with greater hazard of incident pneumonia (PERR-adjusted hazard ratio=1.82, 95% confidence interval=1.27-2.54), accounting for pretreatment pneumonia rates. Estimates were similar across age and comorbidity subgroups. Similar results were also obtained from propensity score- and inverse probability-weighted models.

CONCLUSION

In a large cohort of older adults in primary care, PPI prescription was associated with greater risk of pneumonia in the second year of treatment. Results were robust across alternative analysis approaches. Controversies about the validity of reported short-term harms of PPIs should not divert attention from potential long-term effects of PPI prescriptions on older adults.

Adjusting for unmeasured confounding in nonrandomized longitudinal studies: a methodological review

OBJECTIVES

Motivated by recent calls to use electronic health records for research, we reviewed the application and development of methods for addressing the bias from unmeasured confounding in longitudinal data.

STUDY DESIGN AND SETTING

Methodological review of existing literature. We searched MEDLINE and EMBASE for articles addressing the threat to causal inference from unmeasured confounding in nonrandomized longitudinal health data through quasi-experimental analysis.

RESULTS

Among the 121 studies included for review, 84 used instrumental variable analysis (IVA), of which 36 used lagged or historical instruments. Difference-in-differences (DiD) and fixed effects (FE) models were found in 29 studies. Five of these combined IVA with DiD or FE to try to mitigate for time-dependent confounding. Other less frequently used methods included prior event rate ratio adjustment, regression discontinuity nested within pre-post studies, propensity score calibration, perturbation analysis, and negative control outcomes.

CONCLUSION

Well-established econometric methods such as DiD and IVA are commonly used to address unmeasured confounding in nonrandomized longitudinal studies, but researchers often fail to take full advantage of available longitudinal information. A range of promising new methods have been developed, but further studies are needed to understand their relative performance in different contexts before they can be recommended for widespread use.

Prior event rate ratio adjustment for hidden confounding in observational studies of treatment effectiveness: a pairwise Cox likelihood approach

Observational studies provide a rich source of information for assessing effectiveness of treatment interventions in many situations where it is not ethical or practical to perform randomized controlled trials. However, such studies are prone to bias from hidden (unmeasured) confounding. A promising approach to identifying and reducing the impact of unmeasured confounding is prior event rate ratio (PERR) adjustment, a quasi-experimental analytic method proposed in the context of electronic medical record database studies. In this paper, we present a statistical framework for using a pairwise approach to PERR adjustment that removes bias inherent in the original PERR method. A flexible pairwise Cox likelihood function is derived and used to demonstrate the consistency of the simple and convenient alternative PERR (PERR-ALT) estimator. We show how to estimate standard errors and confidence intervals for treatment effect estimates based on the observed information and provide R code to illustrate how to implement the method. Assumptions required for the pairwise approach (as well as PERR) are clarified, and the consequences of model misspecification are explored. Our results confirm the need for researchers to consider carefully the suitability of the method in the context of each problem. Extensions of the pairwise likelihood to more complex designs involving time-varying covariates or more than two periods are considered. We illustrate the application of the method using data from a longitudinal cohort study of enzyme replacement therapy for lysosomal storage disorders. © 2016 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.