Pharmacoepidemiological research using French reimbursement databases: yes we can!

[Post-marketing surveillance systems for psychoactive prescription drug abuse]

Drugs affecting the central nervous system form a unique group of products for surveillance because they could be misused, abused or diverted. Considering the characteristics of this behaviour that is often concealed, specific post-marketing surveillance systems have been developed to monitor abuse of prescription drugs in some countries. The purpose of this review is to list and to describe post-marketing surveillance systems, according their methodology, in France and in foreign countries. These programs are based on adverse effect notifications, medical or legal consequences of abuse, general or specific population-based survey, professional networks or medication databases. Some programs use simultaneously several information sources. In conclusion, the multifaceted nature, the diversity and the inventiveness of post-marketing surveillance systems reflects the complexity of the abuse issue.

Transition in care in persons with antidepressant prescription in naturalistic conditions

BACKGROUND

Few studies have explored transition to psychiatric care in persons treated by antidepressants in primary care.

OBJECTIVE

To assess the rates of transition in care after a dispensing of antidepressant and the characteristics associated with transition from primary to psychiatric care.

METHODS

A follow-up study of persons (n = 19 975) starting antidepressant treatment was carried out on a representative sample of the 2005-07 French Social Security Insurance national database. Prescribers were categorized as GPs, hospital practitioners (as a proxy for public psychiatrist), private psychiatrists and other private specialists. Transition in care was defined as a prescription of psychotropic drugs by a prescriber belonging to a category different from that of the index antidepressant prescriber.

RESULTS

Eighty per cent of incident antidepressant treatments were initiated by a GP and 28% of persons transited in care. The most frequent pattern was transition from a GP to a hospital practitioner (8%) or to a private psychiatrist (5%). Transition to psychiatric care was independently associated with younger age and markers of illness severity (psychiatric chronic illness status, duration of the index antidepressant treatment >6 months and prescription of other classes of psychotropic drugs). Almost all treatments with antipsychotics or mood stabilizers were initiated after the transition to a specialist.

CONCLUSIONS

Transition to psychiatric care seemed coherent with distribution of tasks between primary and secondary care according to illness severity. Further studies using detailed clinical information are required to assess not only the clinical appropriateness of psychiatric referral but also lack of referral in antidepressant users.

Academic impact of a public electronic health database: bibliometric analysis of studies using the general practice research database

BACKGROUND

Studies that use electronic health databases as research material are getting popular but the influence of a single electronic health database had not been well investigated yet. The United Kingdom's General Practice Research Database (GPRD) is one of the few electronic health databases publicly available to academic researchers. This study analyzed studies that used GPRD to demonstrate the scientific production and academic impact by a single public health database.

METHODOLOGY AND FINDINGS

A total of 749 studies published between 1995 and 2009 with 'General Practice Research Database' as their topics, defined as GPRD studies, were extracted from Web of Science. By the end of 2009, the GPRD had attracted 1251 authors from 22 countries and been used extensively in 749 studies published in 193 journals across 58 study fields. Each GPRD study was cited 2.7 times by successive studies. Moreover, the total number of GPRD studies increased rapidly, and it is expected to reach 1500 by 2015, twice the number accumulated till the end of 2009. Since 17 of the most prolific authors (1.4% of all authors) contributed nearly half (47.9%) of GPRD studies, success in conducting GPRD studies may accumulate. The GPRD was used mainly in, but not limited to, the three study fields of "Pharmacology and Pharmacy", "General and Internal Medicine", and "Public, Environmental and Occupational Health". The UK and United States were the two most active regions of GPRD studies. One-third of GRPD studies were internationally co-authored.

CONCLUSIONS

A public electronic health database such as the GPRD will promote scientific production in many ways. Data owners of electronic health databases at a national level should consider how to reduce access barriers and to make data more available for research.

Nationwide register-based surveillance system on drugs and pregnancy in Finland 1996-2006

PURPOSE

The aim of this population-based nationwide drugs and pregnancy surveillance system was to get basic information on the use of prescribed drugs during pregnancy and to achieve more detailed information on drug exposure-outcome associations with data obtained from the Finnish national health registers.

METHODS

The data are based on information on all pregnancies ending in births (n = 632,629) or termination of pregnancy (TOP) (n = 117,255) in Finland between 1996 and 2006. The data containing information on maternal background factors, chronic diseases and drug purchases during pregnancy, born children, induced abortions, perinatal health and major congenital malformations were obtained from the Finnish national health registers. Information from the different registers was merged through record linkages based on unique personal identification numbers. Statistical analyses were conducted between the exposed and unexposed for individual drugs or Anatomical Therapeutic Chemical (ATC) classification drug groups in multivariate logistic regression including potential confounding factors.

RESULTS

Nearly half of parturients and almost every third woman with TOP purchased drugs at least once 1 month prior to pregnancy and/or during pregnancy. The most frequent chronic diseases during pregnancy were asthma, hypothyroidism, epilepsy, rheumatoid arthritis and diabetes. Previously known causal connections between maternal chronic diseases and/or medication and perinatal health risks were already seen in the ATC-group-level analysis.

CONCLUSIONS

Comprehensive and detailed information for pharmacoepidemiological research on the effects of drug use during pregnancy in a cohort setting is possible with long-term and cumulative data collection.

Patterns of methylphenidate use and assessment of its abuse and diversion in two French administrative areas using a proxy of deviant behaviour determined from a reimbursement database: main trends from 2005 to 2008

BACKGROUND

Methylphenidate is a psychostimulant drug indicated for the treatment of attention-deficit hyperactivity disorder (ADHD). Its abuse and diversion have been previously described in specific populations, such as students; however, few studies investigating abuse and diversion among the overall population are available.

OBJECTIVES

The aim of this study was to describe patterns of methylphenidate use and to explore the magnitude of its abuse and diversion in two French administrative areas using data from a reimbursement database. A proxy of 'deviant behaviour' was used for the abuse and diversion of methylphenidate, defined using the following parameters: total number of defined daily doses (DDDs) of methylphenidate dispensed; number of different pharmacies seen for dispensing of methylphenidate; number of prescribers consulted for a prescription of methylphenidate; and number of dispensings of methylphenidate. Data from the reimbursement database were analysed by clustering methods. These data were assessed from 2005 to 2008.

METHOD

The French General Health Insurance System (GHIS) database was used to obtain data on methylphenidate use in two French administrative areas. Individuals affiliated to the GHIS who had a prescription for methylphenidate reimbursed between 1 January and 31 March of 4 selected years (2005, 2006, 2007 and 2008) were included. After the first dispensing of methylphenidate for these individuals, all their dispensings (including methylphenidate and other psychoactive drugs) were monitored over a 9-month period. Following a descriptive analysis, a clustering method was used to identify different subgroups of subjects according to the methylphenidate consumer profile characteristics.

RESULTS

With regard to the number of patients who had a dispensing for methylphenidate during the first quarter of the year, an 84% increase was observed between 2005 (n = 640) and 2008 (n = 1175). The clustering method identified two subgroups. One of them was characterized by a higher number of dispensings, different prescribers and pharmacies and a greater total dispensed quantity, suggesting a deviant behaviour and, thus, possible abuse and diversion of methylphenidate. These subjects were older (aged 35.4 ± 11.3 years) and were more frequently patients receiving benzodiazepines, antidepressants, antipsychotics and maintenance opioid treatment. The proportion of subjects with a deviant behaviour increased from 0.5% in 2005 to 2% in 2007 and then decreased to 1.2% in 2008.

CONCLUSION

This method was able to assess the magnitude of methylphenidate abuse liability and to follow its evolution. The decrease in methylphenidate abuse and diversion seen between 2007 and 2008 can be explained by the enactment in April 2008 of specific regulations for prescription drugs (such as methylphenidate) that are deemed by the French government to have the potential for misuse; these regulations require the establishment of a 'contract of care' between the GHIS, prescriber and patient.

HPV vaccination in France: uptake, costs and issues for the National Health Insurance

INTRODUCTION

Two vaccines for primary prevention of cervical cancer are available in France, Gardasil® and Cervarix®, since 2007 and 2008 respectively. Currently, the French guidelines indicate vaccination of girls aged 14 with a catch-up program for females from 15 to 23 years old. In France, the reimbursement rate for these vaccines is 65% of the vaccine price, resulting in Gardasil® being the fifth highest drug expenditure of the main scheme of the French National Health Insurance in 2008. The purpose of this study is to provide data on vaccination coverage and costs in France until 31 December 2009. In addition, the current vaccination coverage rate is compared with the coverage rates assumed in cost-effectiveness studies.

METHODS

Data were extracted from the National Health Insurance Information System (SNIIRAM). The SNIIRAM records all reimbursements of medical costs to patients--including drugs--by the French public Health Insurance Schemes since 2004. The analysis was performed for the period of July 2007 until December 2009 using the data of the general scheme of National Health Insurance covering about 88% of the French population, i.e., 56.5 million people. Vaccination rates for one or three doses were determined for the target and catch-up population using the 2009 reference population from the general health insurance scheme as the denominator.

RESULTS

The cumulative number of doses reached 2,900,000 at the end of 2009. About 1,200,000 girls and young women have been reimbursed for at least one vaccine dose, of these 96.5% females aged 14-23 years. Among the target group, reimbursement for at least one dose remained low, from 50.8% for girls aged 14 years in 2007 to 41.7% and 20.5% for girls aged 14 years in 2008 and 2009 respectively. In terms of complete vaccination, only 33.3% of girls of the age of 14 years in 2007 and 23.7% in 2008 were reimbursed for 3 doses of HPV vaccine. The maximum uptake in the catch-up group for both 1 and 3 doses was observed for women born in 1992 (15 years in 2007) with 52.5% and 35.6% respectively.

CONCLUSION

Low rates of coverage have been observed both in the target and catch-up groups in France. Considering this, the cost-effectiveness of vaccination in combination with opportunistic screening or organized screening needs to be re-evaluated.

Which indicators can public health authorities use to monitor prescription drug abuse and evaluate the impact of regulatory measures? Controlling High Dosage Buprenorphine abuse

BACKGROUND

Two methods have been recently developed from a drug reimbursement database to provide useful indicators for public health authorities concerning the abuse potential of psychotropic drugs. The doctor-shopping indicator (DSI) measures the proportion of the drug obtained by doctor shopping among the overall quantity of the drug reimbursed and the clustering method reveals subgroups of deviant patients.

OBJECTIVE

The objective of the study was to analyze and compare indicators resulting from these two methods, applied to High Dosage Buprenorphine (HDB) (a product well-known to be diverted in France), in order to determine which public health authorities needs they answer.

DATA ANALYSIS

The patients with reimbursed HDB were grouped using the clustering method in terms of drug dispensations characteristics over a nine month period. The characteristics of the resulting subgroups, including their DSI, were then compared.

RESULTS

4787 Patients (73.4%) had no measurable doctor-shopping behaviour. But the comparison of the two methods demonstrated that the more a patient's profile was characterized by deviant behavior, the higher was the DSI: from 0.4% in a subgroup with a median profile to 72% in a subgroup with a deviant profile.

CONCLUSION

These two methods are useful surveillance tools for public health authorities: the clustering method may help devise pertinent intervention strategies to reduce prescription drug abuse while the DSI method provides quantitative information demonstrating whether these strategies are useful. We discuss the advantages and disadvantages of using these two methods as useful indicators for public health authorities.

Benfluorex and valvular heart disease: a cohort study of a million people with diabetes mellitus

PURPOSE

To evaluate and quantify in diabetic patients treated with benfluorex in France, a fenfluramine-derivated product, a possible increase in risk of valvular heart disease, previously suggested by several published case reports.

METHODS

This was a French comparative cohort study using data from two large national linked databases, health insurance system (SNIIRAM) and hospitalization (PMSI). Patients aged 40-69 years with reimbursement for oral antidiabetic and/or insulin in 2006 were eligible. Exposed patients were defined as patients with at least one benfluorex reimbursement in 2006. Selected admission diagnoses of interest in 2007 and 2008 PMSI databases were valvular insufficiency for any cause, mitral insufficiency, aortic insufficiency, and valvular replacement surgery with cardiopulmonary bypass. Relative risks (RR) were adjusted on gender, age, and history of chronic cardiovascular disease.

RESULTS

A total of 1,048173 diabetic patients were included, with 43,044 (4.1%) exposed to benfluorex. The risk of hospitalization in 2007 and 2008 for any cardiac valvular insufficiency was higher in the benfluorex group: crude RR=2.9 [95% confidence interval 2.2-3.7] and adjusted RR=3.1 [2.4-4.0], with a lower risk for patients with lower cumulative dose of benfluorex. Adjusted RR for mitral insufficiency and aortic insufficiency admissions were 2.5 [1.9-3.7] and 4.4 [3.0-6.6], respectively. Adjusted RR for valvular replacement surgery was 3.9 [2.6-6.1].

CONCLUSIONS

Benfluorex in diabetic patients was significantly associated with hospitalization for valvular heart disease in the 2 years following benfluorex exposure. Linkage between SNIIRAM and PMSI databases is in France a valuable tool to quantify the risk of serious adverse drug reactions.

Has reimbursement of bone mineral density testing and anti-osteoporotic treatments improved management of osteoporosis in France?

BACKGROUND

Since July 1, 2006 bone mineral density (BMD) testing has been reimbursed for osteoporosis screening under specific clinical conditions in France. Concurrently, drug coverage has been extended to all anti-osteoporotic treatments in women with low BMD. It was thus expected that the utilisation of BMD testing and subsequent treatment would increase significantly.

OBJECTIVE

To examine the pattern of use of BMD testing following reimbursement by the French National Health Insurance and subsequent prescriptions reimbursed for anti-osteoporotic treatment in the general population of the Rhone-Alpes area, France.

METHODS

Time trends, age and sex distribution of the population, and specialty of the physicians who performed the test were examined from all physicians' claims for BMD testing between 2006 and 2009 in Rhone-Alpes. A prospective inception cohort of all women aged 50 years and older, without prior reimbursement for anti-osteoporotic treatment and who had a BMD test between January 1, 2007 and March 31, 2007, was then created. Women were followed for 1 year for their reimbursements of estrogen or other anti-osteoporotic treatment (bisphosphonates, raloxifene, teriparatide and strontium ranelate) after their initial BMD test.

RESULTS

Between July 2006 and August 2009, 140,038 BMD tests were reimbursed in Rhone-Alpes, France, among which 94.8% were performed in women aged 50 years and older. In women, the number of BMD tests tended to significantly decrease during the study period (p for trend=0.047). The inception cohort was composed of 9635 women without prior anti-osteoporotic treatment. Among women without prior estrogen treatment, 27.3% (n=2339) initiated an anti-osteoporotic treatment within 4 months after a BMD test, mostly with bisphosphonates (16.8% n=1441), followed by strontium ranelate (5.8% n=498) and raloxifene (3.4% n=290). At 1 year 1702 women (71.7%) were still on treatment.

CONCLUSION

Following the decision by the French National Health Insurance of repayment of BMD measurement, the proportion of women who had BMD testing tended to decrease instead of increase over the time period. Although BMD screening is focused in France on high risk women, only 25% of those who had an initial BMD test were subsequently treated with an anti-osteoporotic medication.

French national health insurance information system and the permanent beneficiaries sample

In France, in the early 2000s, legislators ordered that the National Health Insurance regime develop an inter-regime information system (SNIIR-AM) aimed at better understanding and evaluating beneficiaries' health care consumption and associated expenditures. In 2009, it contained data from the general health insurance regime that covers 86% of the French population; approximately 53 million people. Data are only available for a period of two years plus the current year. In addition, a permanent sample of health insurance beneficiaries (EGB) was created from the SNIIR-AM database. This is a permanent, representative cross-sectional sample of the population covered by National Health Insurance which, since 2004, monitors beneficiaries' health care consumption over a period of 20 years. It contains anonymous sociodemographic and medical characteristics and records of health care reimbursements. It was created using a systematic sampling method (1/97) on the two-digit control key of beneficiaries' national identification number and includes both current year reimbursement recipients and non-recipients. In 2009, it grouped together almost 500,000 beneficiaries covered by the National Health Insurance Fund for Salaried Workers; 77% of the population residing in France excluding public service employees and students. The EGB is used to conduct longitudinal studies as it permits tracing back patients' care paths and use of care in both hospital and office-based care environments and to calculate individual expenditures. It also permits the study of certain relatively frequent diseases characterised by a 100% reimbursement rate for certain chronic diseases and the reimbursement of tracer drugs. Eventually, the SNIIR-AM will include beneficiaries covered by all the different Health Insurance regimes in France.