New opportunities for drug outcomes research in cancer patients: The linkage of the Eindhoven Cancer Registry and the PHARMO Record Linkage System

https://doi.org/10.1016/j.ejca.2009.09.010Get rights and content

Abstract

Background

Insight into co-morbidity and treatment effects is pivotal to improve quality of care for cancer patients.

Objectives

To determine whether linkage of the Eindhoven Cancer Registry (ECR) and the PHARMO Record Linkage System (RLS) was technically feasible and to assess which patient-centric data would result from this linkage.

Methods

The ECR records data on tumour stage and primary treatment of all newly diagnosed cancer patients in the southeastern Netherlands including co-morbidity at diagnosis, whereas the PHARMO RLS includes data from multiple linked observational databases such as data on drug utilisation (for both in- and out-patients, including chemotherapy), hospitalisations and clinical laboratory measurements. All patients who lived or had been living in the overlapping area served by the ECR and the PHARMO RLS during 1998–2006 were selected for linkage which was performed with probabilistic medical record linkage.

Results

The linkage resulted in an ECR-PHARMO cohort of 40,004 cancer patients with a total of 42,767 primary tumours. The cancer patients in the linked ECR-PHARMO cohort were representatives for the cancer patients included in the total ECR during 1998–2006. Cancer patients included in the cohorts had a mean history of 5 years and a mean follow-up ranging from 2 to more than 4 years (dependent on the survival rate of the specific cancer type).

Conclusions

Linkage of ECR and the PHARMO RLS creates the possibility to study patient-centric drug utilisation, health resources utilisation and their costs, in addition to the effectiveness and safety of pharmaceuticals in routine daily practice in cancer patients.

Introduction

In Europe there are over two million incident cases of cancer every year and it is expected that this number will continue to rise.1 With the ageing of the population and the increasing survival of cancer patients with co-morbid conditions, cancer is and will become an increasingly important factor in the global burden of disease in the decades to come.

A growing number of cancer patients are treated with novel treatment combinations and targeted therapies are being introduced at relatively early stages of disease progression. While effectiveness and safety of these new therapies are thoroughly studied in randomized clinical trials, little data are available on these parameters in a daily practice setting. Effectiveness of anti-cancer drugs may be different than expected as daily clinical practice differs from the experimental setting with respect to the heterogeneity of patients, their treatments and co-medication. Moreover, important safety issues may not be detectable in clinical trials since the frequency of many adverse events is low and adverse effects of cancer therapies may occur many years after drug administration.2, 3, 4 Balancing effectiveness and adverse effects is therefore a major challenge in the treatment of cancer patients and monitoring post approval drug use is especially important for cancer drugs that receive accelerated FDA and EMEA approval.

Besides monitoring the effectiveness and safety of cancer treatments, pharmaceuticals used for a wide variety of other conditions may be associated with increasing or decreasing risk of cancer, such as non-steroidal anti-inflammatory drugs or lipid-lowering drugs.5, 6 Data to confirm, refute or elaborate on these hypotheses are sparse.

To shed light on the burden of cancer, disease management, safety aspects, co-morbidity patterns and outcomes assessments, follow-up of patients before and after cancer diagnosis in routine daily practice is pivotal. For this, large administrative databases or cancer registries are often used.7 However, both type of data collections mostly lack detail either with respect to the incidence and staging of cancer or with respect to (pharmaco)treatments, morbidity and co-morbidity during follow-up.8

We therefore explored whether the linkage of a regional cancer disease register (Eindhoven Cancer Registry: ECR) and a patient-centric data network including multiple linked observational databases (PHARMO Record Linkage System: PHARMO RLS9) was feasible, valid and detailed enough to fulfil the abovementioned information needed. In this manuscript we describe the results of this linkage as well as the patient-centric data on drug treatment (out-patient and in-patient), hospitalisations and clinical laboratory measurements that become available during follow-up before and after cancer diagnosis for the cancer patients present in these linked databases.

Section snippets

Data sources

The Eindhoven Cancer Registry (ECR) is a population-based registry (covering a demographic region with 2.4 million inhabitants) which is maintained by the Comprehensive Cancer Centre South and collates records on all newly diagnosed cancer patients in the southeastern part of the Netherlands.10, 11 The ECR is notified for new cases of cancer by 6 pathology departments, 10 general hospitals and two radiotherapy institutes. Trained registry personnel subsequently actively collect on site data on

Results

Of the 104,562 cancer patients registered in the ECR in the period 1998–2006, 57,612 (55%) patients were not eligible for linkage of which 2339 (2%) were diagnosed in hospitals elsewhere in the Netherlands and 55,273 (53%) were not living in the ECR-PHARMO catchment area (Fig. 1). Of the remaining 46,950 patients who were eligible for linkage, 40,004 (85%) cancer patients were finally linked and included in the ECR-PHARMO cohort with a total of 42,767 primary tumours. Detailed tumour data and

Discussion

In this study, the ECR and the PHARMO RLS were linked creating a novel ECR-PHARMO population-based and patient-centric cohort that provides an excellent opportunity to gain real-life insights into drug utilisation (for both in- and out-patients), health resources utilisation and their costs. Moreover, the effectiveness and safety of pharmaceuticals in cancer patients outside the clinical trial setting can be assessed using hospitalisation data and clinical laboratory measurements. This linkage

Conflict of interest statement

No conflict of interest declared for the authors Dr. L.V. van de Poll-Franse, Dr. V.E.P.P. Lemmens, Dr. G. Vreugdenhil, Dr. J.F.M. Pruijt and Prof. Dr. J.W.W. Coebergh. Mrs. M.P.P. van Herk-Sukel and Dr. R.M.C. Herings are employees of the PHARMO Institute for Drug Outcomes Research. This research institute performs financially supported studies for several pharmaceutical companies. This study however, was performed as part of the PhD programme of Mrs. M.P.P. van Herk-Sukel, initiated by Dr.

Acknowledgements

The authors would like to thank the participating pharmacies for the opportunity to validate ECR-PHARMO linkage. The authors thank Mark AGM Tinga, PhD [employee of the PHARMO Institute for Drug Outcomes Research], for performing the medical record linkage. This work was partially supported by the Netherlands organization for health research and development [ZonMw 80-82500-98-8227].

References (25)

  • Y.S. Zhao et al.

    Association between NSAIDs use and breast cancer risk: a systematic review and meta-analysis

    Breast Cancer Res Treat

    (2009)
  • J.Z. Ayanian et al.

    Enhancing research on cancer survivors

    J Clin Oncol

    (2006)
  • Cited by (104)

    • Real-world data of HER2-low metastatic breast cancer: A population based cohort study

      2022, Breast
      Citation Excerpt :

      This population-based network of electronic healthcare databases combines data from different primary and secondary healthcare settings in the Netherlands [17]. These different data sources are linked on a patient level through validated algorithms [18]. A detailed description of the databases used for this study is included in previous published papers [16,17].

    View all citing articles on Scopus
    View full text