Validation and application of a death proxy in adult cancer patients

Impact of coordinated care on adherence to antihypertensive medicines among adults experiencing polypharmacy in Australia

BACKGROUND

Adherence to antihypertensives is key for blood pressure control. Most people with hypertension have several comorbidities and require multiple medicines, leading to complex care pathways. Strategies for coordinating medicine use can improve adherence, but cumulative benefits of multiple strategies are unknown.

METHODS

Using dispensing claims for a 10% sample of eligible Australians, we identified adult users of antihypertensives during July 2018-June 2019 who experienced polypharmacy (≥5 unique medicines). We measured medicine use reflecting coordinated medicine management in 3 months before and including first observed dispensing, including: use of simple regimens for each cardiovascular medicine; prescriber continuity; and coordination of dispensings at the pharmacy. We measured adherence (proportion of days covered) to antihypertensive medicines in the following 12 months, and used logistic regression to assess independent associations and interactions of adherence with these measures of care.

RESULTS

We identified 202 708 people, of which two-thirds (66.6%) had simple cardiovascular medicine regimens (one tablet per day for each medicine), two-thirds (63.3%) were prescribed >75% of medicines from the same prescriber, and two-thirds (65.5%) filled >50% of their medicine on the same day. One-third (28.4%) of people experienced all three measures of coordinated care. Although all measures were significantly associated with higher adherence, adherence was greatest among people experiencing all three measures (odds ratio = 1.63; 95% confidence interval: 1.55-1.72). This interaction was driven primarily by effects of prescriber continuity and dispensing coordination.

CONCLUSIONS

Coordinating both prescribing and dispensing of medicines can improve adherence to antihypertensives, which supports strategies consolidating both prescribing and supply of patients' medicines.

Validation study on definition of cause of death in Japanese claims data

Identifying the cause of death is important for the study of end-of-life patients using claims data in Japan. However, the validity of how cause of death is identified using claims data remains unknown. Therefore, this study aimed to verify the validity of the method used to identify the cause of death based on Japanese claims data. Our study population included patients who died at two institutions between January 1, 2018 and December 31, 2019. Claims data consisted of medical data and Diagnosis Procedure Combination (DPC) data, and five definitions developed from disease classification in each dataset were compared with death certificates. Nine causes of death, including cancer, were included in the study. The definition with the highest positive predictive values (PPVs) and sensitivities in this study was the combination of "main disease" in both medical and DPC data. For cancer, these definitions had PPVs and sensitivities of > 90%. For heart disease, these definitions had PPVs of > 50% and sensitivities of > 70%. For cerebrovascular disease, these definitions had PPVs of > 80% and sensitivities of> 70%. For other causes of death, PPVs and sensitivities were < 50% for most definitions. Based on these results, we recommend definitions with a combination of "main disease" in both medical and DPC data for cancer and cerebrovascular disease. However, a clear argument cannot be made for other causes of death because of the small sample size. Therefore, the results of this study can be used with confidence for cancer and cerebrovascular disease but should be used with caution for other causes of death.

Handgrip strength and respiratory disease mortality: Longitudinal analyses from SHARE

BACKGROUND

While the association between handgrip strength and all-cause mortality is more deeply explored, no previous studies have been specifically focused on handgrip strength and respiratory disease mortality. The purpose of the study was to investigate the association between handgrip strength and respiratory disease mortality in a large representative sample.

METHODS

Individuals aged 50 or over from 27 European countries and Israel participated in this longitudinal study. Data on handgrip strength and all-cause and respiratory disease mortality were retrieved from the Survey of Health, Ageing and Retirement in Europe (SHARE) waves 1, 2, 4, 5, 6 and 7. We estimated the sub hazard ratios (SHRs) for respiratory disease mortality using a Fine-Gray sub-distribution method with both time-varying exposure and covariates and mortality due to other causes as competing risk. Furthermore, we assessed dose-response associations of handgrip strength (modelled as a continuous exposure) with respiratory disease mortality using restricted cubic splines and estimated hazard ratios (HRs).

RESULTS

We included 60,883 men and 74,904 women with a mean age of 63.6 (SD 9.7) years at study entry. During a median (interquartile range) of 7.4 years of follow-up 565 (0.4%) participants died due to respiratory diseases. The increase of 1 single kg of handgrip strength showed a 6% incidence reduction on respiratory disease mortality (SHR, 0.94; 95%CI, 0.92-0.96) after adjusting for potential confounders. Furthermore, each kg increase of handgrip strength reduced respiratory disease mortality risk in a dose-response fashion and a significant threshold for values of 41 kg (HR, 0.49; 95%CI, 0.26-0.92) and higher was identified.

CONCLUSIONS

Higher handgrip strength is associated with lower mortality due to respiratory disease. Intervention studies are needed to determine whether strength training in respiratory disease patients can prevent premature mortality.

Longitudinal association of handgrip strength with all-cause and cardiovascular mortality in older adults using a causal framework

To date, there is no study addressing the time-varying confounding bias in the association of handgrip strength (HGS) with all-cause or cardiovascular mortality. Therefore, we conducted marginal structural models (MSM) to provide causal estimations on the associations of HGS with all-cause and cardiovascular mortality in a representative sample of adults aged 50 years or older. Data from 29 countries including 121,116 participants (276,994 observations; mean age 63.7 years; 56.3 % women) free from prior heart attack or stroke were retrieved from consecutive waves of the Survey of Health, Ageing and Retirement in Europe (SHARE). During a median of 7.7 years follow-up (interquartile range 3.8-11.8) and 1,009,862 person-years, 6407 participants (5.3 %) died due to all causes, and 2263 (1.9 %) died due to cardiovascular diseases. Using repeated measures of handheld dynamometry, we determined absolute and relative to body mass index HGS of each participant. We applied adjusted MSM to estimate hazard ratios (HRs) associated with changes over time in HGS addressing the time-varying confounding bias. An increase of 5 kg in HGS was associated with a reduced risk of all-cause [HR 0.86, 95 % confidence interval (CI), 0.86-0.90], overall cardiovascular (HR 0.86, 95 % CI 0.82-0.86), heart attack (HR 0.90, 95 % CI 0.86-0.95), and stroke (HR 0.86, 95 % CI 0.82-0.90) mortality. The associations of relative HGS were of stronger magnitude in all cases. Our findings provide critical evidence on the importance of increasing general muscle strength in older adults to reduce mortality risk, particularly concerning cardiovascular causes.

Optimal Indicator of Death for Using Real-World Cancer Patients' Data From the Healthcare System

Background: Information on patient’s death is a major outcome of health-related research, but it is not always available in claim-based databases. Herein, we suggested the operational definition of death as an optimal indicator of real death and aim to examine its validity and application in patients with cancer.

Materials and methods: Data of newly diagnosed patients with cancer between 2006 and 2015 from the Korean National Health Insurance Service—National Sample Cohort data were used. Death indicators were operationally defined as follows: 1) in-hospital death (the result of treatment or disease diagnosis code from claims data), or 2) case wherein there are no claims within 365 days of the last claim. We estimated true-positive rates (TPR) and false-positive rates (FPR) for real death and operational definition of death in patients with high-, middle-, and low-mortality cancers. Kaplan−Meier survival curves and log-rank tests were conducted to determine whether real death and operational definition of death rates were consistent.

Results: A total of 40,970 patients with cancer were recruited for this study. Among them, 12,604 patients were officially reported as dead. These patients were stratified into high- (lung, liver, and pancreatic), middle- (stomach, skin, and kidney), and low- (thyroid) mortality groups consisting of 6,626 (death: 4,287), 7,282 (1,858), and 6,316 (93) patients, respectively. The TPR was 97.08% and the FPR was 0.98% in the high mortality group. In the case of the middle and low mortality groups, the TPR (FPR) was 95.86% (1.77%) and 97.85% (0.58%), respectively. The overall TPR and FPR were 96.68 and 1.27%. There was no significant difference between the real and operational definition of death in the log-rank test for all types of cancers except for thyroid cancer.

Conclusion: Defining deaths operationally using in-hospital death data and periods after the last claim is a robust alternative to identifying mortality in patients with cancer. This optimal indicator of death will promote research using claim-based data lacking death information.

Associations of handgrip strength with all-cause and cancer mortality in older adults: a prospective cohort study in 28 countries

Background

mixed evidence exists on the association between muscle strength and mortality in older adults, in particular for cancer mortality.

Aim

to examine the dose–response association of objectively handgrip strength with all-cause and cancer mortality.

Study Design and Setting

data from consecutive waves from the Survey of Health, Ageing and Retirement in Europe comprising 27 European countries and Israel were retrieved. Overall, 54,807 men (45.2%; 128,753 observations) and 66,576 women (54.8%; 159,591 observations) aged 64.0 (SD 9.6) and 63.9 (SD 10.2) years, respectively, were included. Cox regression and Fine-Grey sub-distribution method were conducted.

Results

during the follow-up period (896,836 person-year), the fully adjusted model showed the lowest significant risk estimates for the highest third of handgrip strength when compared with the first third (reference) in men (hazard ratio [HR], 0.41; 95% confidence interval [CI], 0.34–0.50) and women (HR, 0.38; 95% CI, 0.30–0.49) for all-cause mortality. We identified a maximal threshold for reducing the risk of all-cause mortality for men (42 kg) and women (25 kg), as well as a linear dose–response association in participants aged 65 or over. No robust association for cancer mortality was observed.

Conclusion

these results indicate an inverse dose–response association between incremental levels of handgrip and all-cause mortality in older adults up to 42 kg for men and 25 kg for women, and a full linear association for participants aged 65 years or over. These findings warrant preventive strategies for older adults with low levels of handgrip strength.

Trajectories of pregabalin use and their association with longitudinal changes in opioid and benzodiazepine use

ABSTRACT

Concomitant use of pregabalin with opioids and/or benzodiazepines is common, despite the increased risks. However, clinical trials suggest pregabalin can have an opioid-sparing effect when treating acute postoperative pain. We explored how opioid and benzodiazepine use changed over time in people initiating pregabalin, using dispensing claims data for a 10% sample of Australians (2013-19). Among 142,776 people initiating pregabalin (median age = 61 years, 57% female), we used group-based trajectory modelling to identify 6 pregabalin dose trajectories in the first year postinitiation. Two trajectories involved discontinuation: after one dispensing (49%), and after 6 months of treatment (14%). Four trajectories involved persistent use with variable estimated median daily doses of 39 mg (16%), 127 mg (14%), 276 mg (5%), and 541 mg (2%). We quantified opioid and benzodiazepine use in the year before and after pregabalin initiation using generalised linear models. Over the study period, 71% were dispensed opioids and 34% benzodiazepines, with people on the highest pregabalin dose having highest rates of use. Opioid use increased postpregabalin initiation. Among people using both opioids and pregabalin, the geometric mean daily dose in oral morphine equivalents increased after pregabalin initiation in all trajectories, ranging from +5.9% (99% confidence interval 4.8%-7.0%) to +39.8% (99% confidence interval 38.3%-41.5%) in people on the highest daily pregabalin dose. Among people using both pregabalin and benzodiazepines, the dose remained constant over time for people in all trajectories. Notwithstanding its reputation as opioid-sparing, in this outpatient setting, we observed that people using opioids tended to use higher opioid daily doses after pregabalin initiation, especially those on high pregabalin doses.

Incidence and risk factors of hypertension therapy in Australian cancer patients treated with vascular signalling pathway inhibitors

BACKGROUND

Clinical trials report systemic hypertension is an adverse effect of vascular signalling pathway inhibitor (VSPi) use. There are limited data from routine clinical practice. We aimed to estimate the real-world incidence and risk factors of new-onset and aggravated hypertension for cancer patients dispensed VSPi in whole-of-population Australian setting.

METHODS

We used dispensing records for a 10% random sample of Australians to identify treatment with subsidised VSPi from 2013 to 2018. We further identified dispensings of oral antihypertensive medicines 6 months before and 12 months after VSPi therapy. We defined (i) new-onset hypertension in people first dispensed antihypertensives after VSPi and (ii) aggravated hypertension in people with prior antihypertensive use dispensed an additional, or higher strength, antihypertensive after VSPi. We applied the Fine-Gray cumulative incidence function and Cox proportional hazard regression.

RESULTS

1802 patients were dispensed at least one VSPi. The mean age of the cohort was 65 years and 57% were male. The incidence of new-onset treated hypertension was 24.3% (95%CI: 21.2-27.8); age ≥ 60 years (HR 1.74; 95%CI: 1.32-2.31) and treatment with oral tyrosine kinase inhibitors compared to bevacizumab (HR 1.96; 95%CI: 1.16-3.31) were risk factors. The incidence of aggravated hypertension was 25.2% (95%CI: 22.0-28.7) and risk was elevated for patients with renal cancer (HR 2.84; 95%CI: 1.49-5.41) and cancers other than colorectal (HR 1.85; 95%CI: 1.12-3.03).

CONCLUSIONS

Our real-world estimates of incident hypertension appear comparable to those observed in clinical trials (21.6-23.6%). Our population-based study provides some insight into the burden of hypertension in patients commencing VSPi in routine practice.

Medicine treatment of glaucoma in Australia 2012-2019: prevalence, incidence and persistence

Objective

Medical therapy can halt or significantly slow the progression of glaucoma if medicines are used in accordance with the guidelines. We used dispensing claims for a 10% sample of all Australians dispensed publicly subsidised glaucoma medicines to determine the prevalence and incidence of glaucoma medicine treatment and to examine treatment persistence between July 2012 and June 2019.

Methods

We estimated incidence and prevalence per 10 000 population for Australian financial years (1 July to 30 June). We defined prevalence as at least one dispensing of any glaucoma medicine and incidence as a dispensing of any glaucoma medicine with no previous dispensing during the preceding 12 months. We estimated duration of treatment for a cohort initiating glaucoma medicines and used Kaplan-Meier methods to estimate the proportion of people persisting on treatment at 6, 12, 18 and 36 months after initiation. We stratified analyses by the number of repeats prescribed at initiation, age, sex and medicine class.

Results

Prevalence remained stable over the study period at around 180/10 000 people/year; incidence was also stable around 36/10 000/year. Among 34 900 people initiating glaucoma medicines, 37.0% remained on treatment at 6 months from initiation, 29.8% at 12 months and 19.2% at 36 months. Median duration of treatment was 13.2 months (IQR: 2.5—not reached) for people initiating prostaglandin analogues and less than 3 months for those initiating other medicine classes.

Conclusion

Prevalence and incidence of glaucoma treatment have not changed in Australia over the past decade. Persistence to treatment increased with age but remained poor throughout the study period.

Identifying incident cancer cases in dispensing claims: A validation study using Australia's Repatriation Pharmaceutical Benefits Scheme (PBS) data

Introduction

Dispensing claims are used commonly as proxy measures in pharmacoepidemiological studies; however, their validity is often untested.

Objectives

To assess the performance of a proxy for identifying cancer cases based on the dispensing of anticancer medicines and estimate the misclassification of cancer status and potential for bias researchers may encounter when using this proxy.

Methods

We conducted our validation study using Department of Veterans’ Affairs (DVA) client data linked with the New South Wales (NSW) Cancer Registry and Repatriation Pharmaceutical Benefits Scheme data. We included DVA clients aged ≥65 years residing in NSW between July 2004 and December 2012. We matched clients with a cancer diagnosis to clients without a diagnosis based on demographic characteristics and available observation time. We used dispensing claims for anticancer medicines dispensed between July 2004 and December 2013 as a proxy to identify clients with cancer and calculated sensitivity, specificity, positive predictive values and negative predictive values compared with cancer registrations (gold standard), overall and by cancer site. We illustrated misclassification by the proxy in a cohort of people initiating opioid therapy. Using the proxy, we excluded people with cancer from the cohort, in an attempt to delineate people potentially using opioids for cancer rather than chronic non-cancer pain.

Results

We identified 15,679 new cancer diagnoses in 14,112 DVA clients from the cancer registry and 62,663 clients without a diagnosis. Sensitivity of the proxy based on dispensing claims was 30% for all cancers and around 20% for specific cancers (range: 10-67%). Specificity was above 90% for all cancers. The dispensing proxy correctly identified 26% of people with a cancer diagnosis who initiated opioid therapy and failed to identify 74% those with a cancer diagnosis; the proxy was most robust for clients with breast cancer where 61% were correctly identified by proxy.

Conclusions

Using dispensing of anticancer medicines to identify people with a cancer diagnosis performed poorly. Excluding patients with evidence of anticancer medicine use from cohort studies may result removal of a disproportionate number of women with breast cancer. Researchers excluding or otherwise using anticancer medicine dispensing to identify people with cancer in pharmacoepidemiological studies should acknowledge the potential biases introduced to their findings.

Keywords

cancer, diagnosis, proxy, dispensing records, validation study

Application and comparison of generalized propensity score matching versus pairwise propensity score matching

AIM

A comparison of conventional pairwise propensity score matching (PSM) and generalized PSM method was applied to the comparative effectiveness of multiple treatment options for lung cancer.

MATERIALS & METHODS

Deidentified data were analyzed. Covariate balances between compared treatments were assessed before and after PSM. Cox proportional hazards regression compared overall survival after PSM.

RESULTS & CONCLUSION

The generalized PSM analyses were able to retain 61.2% of patients, while the conventional PSM analyses were able to match from 24.1 to 77.1% of patients from each treatment comparison. The generalized PSM achieved statistical significance (p < 0.05) in 8/10 comparisons, whereas conventional pairwise PSM achieved 1/10. The noted differences arose from different matched patient samples and the size of the samples.

Real-World Evidence: A Comparison of the Australian Herceptin Program and Clinical Trials of Trastuzumab for HER2-Positive Metastatic Breast Cancer

INTRODUCTION

Estimating the real-world cost-effectiveness of a new drug relies on understanding the differences between clinical trial data (pre-reimbursement) and clinical practice (post-reimbursement). This is important for decision makers when reviewing reimbursement decisions, prices, and considering other drugs for the same condition. Differences can arise from differences in patient characteristics, but also from the availability of new evidence and evolving treatment practices. This paper examines these issues using a case study.

METHODS

In 2001, the Australian Government funded trastuzumab for the treatment of HER2+ metastatic breast cancer through the Herceptin Program. The administrative arrangements of the Program resulted in rich observational data that captured information about patients treated with trastuzumab between 2001 and 2010 (n = 3830). The dataset included patient characteristics, dispensed medicines, medical service use and date of death.

RESULTS

Compared to participants in the clinical trials, patients were older, received more prior chemotherapies and a broader range of co-administered chemotherapies. Treatment practices differed from the clinical trials, but also changed over time. For example, in situ hybridization testing, rather than immunohistochemistry testing, and a three weekly administration schedule, rather than one weekly, were increasingly used. Compared to the clinical trials, patients administered trastuzumab with a concomitant chemotherapy generally experienced longer overall survival (151.3 weeks, 95 % CI: 142.6, 163.4), while those who received trastuzumab as a monotherapy experienced shorter overall survival (94.4 weeks, 95%CI: 86.4, 102.9). These findings may be due to a differing relative treatment effect in clinical practice, but may also be due to a range of other factors.

CONCLUSION

This analysis demonstrates the challenges for decision makers that arise because new evidence and evolving treatment practices create a gap between clinical trial data and real-world clinical practice and outcomes.

Does tablet formulation alone improve adherence and persistence: a comparison of ezetimibe fixed dose combination versus ezetimibe separate pill combination?

AIMS

The aim of this study was to compare adherence and persistence in patients who add ezetimibe to statin therapy as a separate pill combination (SPC) or fixed dose combination (FDC).

METHOD

This is a retrospective cohort study of prescription data conducted in an Australian health dataset. Two cohorts were identified: those dispensed statins and subsequently ezetimibe as either SPC or FDC. We compared adherence to combination therapy using the medication possession ratio (MPR), multivariate linear and logistic regression. Persistence to initial combination medicines and any lipid-lowering therapies were analysed using Kaplan Meyer survival and Cox proportional hazards models.

RESULTS

A total of 3651 people initiated ezetimibe SPC and 5740 ezetimibe FDC. There was no significant difference in adherence with mean MPRs: ezetimibe SPC = 0.99 (95% confidence interval 0.98-1.01) and FDC = 0.97 (95% CI 0.95-0.99). One year persistence rates to initial combination medicines were ezetimibe SPC 49.1% vs. FDC 62.4%; hazard ratio (HR) = 1.81 (95% CI 1.76-1.90). However, persistence to any lipid-lowering therapy was higher in those initiating ezetimibe SPC = 84.9% vs. FDC = 76%; HR = 0.62 (95% CI 0.55-0.72). One year persistence rates to any two lipid-lowering medicines were similar: ezetimibe SPC 65.2% and FDC 65%.

CONCLUSION

In this study FDCs have little impact on either adherence or persistence to combination lipid-lowering therapy in people who have been taking statins. The benefit of higher persistence to FDCs in first episode of treatment with initial medicines is debatable as persistence to dual therapy was similar in both cohorts.

Validating a proxy for disease progression in metastatic cancer patients using prescribing and dispensing data

BACKGROUND

Routine data collections are used increasingly to examine outcomes of real-world cancer drug use. These datasets lack clinical details about important endpoints such as disease progression.

AIM

To validate a proxy for disease progression in metastatic cancer patients using prescribing and dispensing claims.

METHODS

We used data from a cohort study of patients undergoing chemotherapy who provided informed consent to the collection of cancer-treatment data from medical records and linkage to pharmaceutical claims. We derived proxy decision rules based on changes to drug treatment in prescription histories (n = 36 patients) and validated the proxy in prescribing data (n = 62 patients). We adapted the decision rules and validated the proxy in dispensing data (n = 109). Our gold standard was disease progression ascertained in patient medical records. Individual progression episodes were the unit of analysis for sensitivity and Positive Predictive Value (PPV) calculations and specificity and Negative Predictive Value (NPV) were calculated at the patient level.

RESULTS

The sensitivity of our proxy in prescribing data was 74.3% (95% Confidence Interval (CI), 55.6-86.6%) and PPV 61.2% (95% CI, 45.0-75.3%); specificity and NPV were 87.8% (95% CI, 73.8-95.9%) and 100% (95% CI, 90.3-100%), respectively. In dispensing data, the sensitivity of our proxy was 64% (95% CI, 55.0-77.0%) and PPV 56.0% (95% CI, 43.0-69.0%); specificity and NPV were 81% (95% CI, 70.05-89.0%) and 91.0% (95% CI, 82.0-97.0%), respectively.

CONCLUSION

Our proxy overestimated episodes of disease progression. The proxy's performance is likely to improve if the date of prescribing is used instead of date of dispensing in claims data and by incorporating medical service claims (such as imaging prior to drug changes) in the algorithm. Our proxy is not sufficiently robust for use in real world comparative effectiveness research for cancer medicines.

Studies using Australia's Pharmaceutical Benefits Scheme data for pharmacoepidemiological research: a systematic review of the published literature (1987-2013)

PURPOSE

Research using dispensing claims is used increasingly to study post-market medicines use and outcomes. The purpose of this review is to catalogue more than 25 years of published literature using Australia's Pharmaceutical Benefits Scheme (PBS) dispensing records.

METHODS

We searched MEDLINE, PreMEDLINE and Embase and conducted author searches for studies published from 1987 to 2013. Independent reviewers screened abstracts of 3209 articles and reviewed 264 full-text manuscripts. Included studies used PBS dispensing data to measure patterns and/or outcomes of prescribed medicines use or dispensing claims to derive a proxy for a specific disease cohort or health outcome.

RESULTS

Of the 228 studies identified, 106 used PBS claims only (56 using claims-level data and 50 using individual-level data) and 63 studies linked individual-level PBS claims to other health data. Most commonly, studies examined trends in drug utilisation (33%), clinician and patient practices (26%), drug use and outcomes (18%) and evaluations of intervention impacts (17%). Sixty-two percent of studies using individual-level data were based on a subset of elderly Australians. Most studies focused on drug classes acting on the nervous system (36%), cardiovascular system (15%) and alimentary tract (11%). Few studies examined prescribed medicines use in children and pregnant women.

CONCLUSIONS

Pharmaceutical Benefits Scheme claims represent a significant resource to examine Australia's billion-dollar annual investment in prescribed medicines. The body of research is growing and has increased in complexity over time. Australia has great potential to undertake world-class, whole-of-population pharmacoepidemiological studies. Recent investment in data linkage infrastructure will significantly enhance these opportunities.

The use and impact of cancer medicines in routine clinical care: methods and observations in a cohort of elderly Australians

INTRODUCTION

After medicines have been subsidised in Australia we know little about their use in routine clinical practice, impact on resource utilisation, effectiveness or safety. Routinely collected administrative health data are available to address these issues in large population-based pharmacoepidemiological studies. By bringing together cross-jurisdictional data collections that link drug exposure to real-world outcomes, this research programme aims to evaluate the use and impact of cancer medicines in a subset of elderly Australians in the real-world clinical setting.

METHODS AND ANALYSIS

This ongoing research programme involves a series of retrospective cohort studies of Australian Government Department of Veterans' Affairs (DVA) clients. The study population includes 104 635 veterans who reside in New South Wales, Australia, and were aged 65 years and over as of 1 July 2004. We will investigate trends in cancer medicines use according to cancer type and other sociodemographic characteristics as well as predictors of the initiation of cancer medicines and other treatment modalities, survival and adverse outcomes among patients with cancer. The programme is underpinned by the linkage of eight health administrative databases under the custodianship of the DVA and the New South Wales Ministry of Health, including cancer notifications, medicines dispensing data, hospitalisation data and health services data. The cancer notifications database is available from 1994 with all other databases available from 2005 onwards.

ETHICS AND DISSEMINATION

Ethics approval has been granted by the DVA and New South Wales Population and Health Service Research Ethics Committees.

RESULTS

Results will be reported in peer-reviewed publications, conference presentations and policy forums. The programme has high translational potential, providing invaluable evidence about cancer medicines in an elderly population who are under-represented in clinical trials.

Claims-based definition of death in Japanese claims database: validity and implications

Background

For the pending National Claims Database in Japan, researchers will not have access to death information in the enrollment files. We developed and evaluated a claims-based definition of death.

Methodology/Principal Findings

We used healthcare claims and enrollment data between January 2005 and August 2009 for 195,193 beneficiaries aged 20 to 74 in 3 private health insurance unions. We developed claims-based definitions of death using discharge or disease status and Charlson comorbidity index (CCI). We calculated sensitivity, specificity and positive predictive values (PPVs) using the enrollment data as a gold standard in the overall population and subgroups divided by demographic and other factors. We also assessed bias and precision in two example studies where an outcome was death. The definition based on the combination of discharge/disease status and CCI provided moderate sensitivity (around 60%) and high specificity (99.99%) and high PPVs (94.8%). In most subgroups, sensitivity of the preferred definition was also around 60% but varied from 28 to 91%. In an example study comparing death rates between two anticancer drug classes, the claims-based definition provided valid and precise hazard ratios (HRs). In another example study comparing two classes of anti-depressants, the HR with the claims-based definition was biased and had lower precision than that with the gold standard definition.

Conclusions/Significance

The claims-based definitions of death developed in this study had high specificity and PPVs while sensitivity was around 60%. The definitions will be useful in future studies when used with attention to the possible fluctuation of sensitivity in some subpopulations.

Validation of a proxy for estrogen receptor status in breast cancer patients using dispensing data

AIM

To assess the performance of a proxy for estrogen receptor (ER) status in breast cancer patients using dispensing data.

METHODS

We derived our proxy using 167 patients. ER+ patients had evidence of at least one dispensing record for hormone therapy during the lookback period, irrespective of diagnosis date and ER- had no dispensing records for hormone therapy during the period. We validated the proxy against our gold standard, ER status from pathology reports or medical records. We assessed the proxy's performance using three lookback periods: 4.5 years, 2 years, 1 year.

RESULTS

More than half of our cohort (62%) were >50 years, 54% had stage III/IV breast cancer at recruitment, (46%) were diagnosed with breast cancer in 2009 and 23% were diagnosed before 2006. Sensitivity and specificity were high for the 4.5 year lookback period (93%, 95% CI: 86-96%; and 95%: 83-99%), respectively) and remained high for the 2-year lookback period (91%: 84-95%; and 95%: 83-99%). Sensitivity decreased (83%: 75.2-89%) but specificity remained high (95%: 83-99%) using the 1-year lookback period and the period is long enough to allow sufficient time for hormone therapy to be dispensed.

CONCLUSION

Our proxy accurately infers ER status in studies of breast cancer treatment based on secondary health data. The proxy is most robust with a minimum lookback period of 2 years.