A Consensus German Reporting Standard for Secondary Data Analyses, Version 2 (STROSA-STandardisierte BerichtsROutine für SekundärdatenAnalysen).

Although secondary data analyses have been established in recent years in health research, explicit recommendations for standardized, transparent and complete reporting of secondary data analyses do not exist as yet. Therefore, between 2009 and 2014, a first proposal for a specific reporting standard for secondary data analysis was developed (STROSA 1). Parallel to this national process in Germany, an international reporting standard for routine data analysis was initiated in 2013 (RECORD). Nevertheless, because of the specific characteristics of the German health care system as well as specific data protection requirements, the need for a specific German reporting standard for secondary data analyses became evident. Therefore, STROSA was revised and tested by a task force of 15 experts from the working group Collection and Use of Secondary Data (AGENS) of the German Society for Social Medicine and Prevention (DGSMP) and the German Society for Epidemiology (DGEpi) as well as from the working group Validation and Linkage of Secondary Data of the German Network for Health Services Research (DNVF). The consensus STROSA-2 checklist includes 27 criteria, which should be met in the reporting of secondary data analysis from Germany. The criteria have been illustrated and clarified with specific explanations and examples of good practice. The STROSA reporting standard aims at stimulating a wider scientific discussion on the practicability and completeness of the checklist. After further discussions and possibly resulting modifications, STROSA shall be implemented as a reporting standard for secondary data analyses from Germany. This will guarantee standardized and complete information on secondary data analyses enabling assessment of their internal and external validity.

[Good Practice of Secondary Data Analysis (GPS): guidelines and recommendations]. / Gute Praxis Sekundärdatenanalyse (GPS): Leitlinien und Empfehlungen.

In 2005, the Working Group for the Survey and Utilisation of Secondary Data (AGENS) of the German Society for Social Medicine and Prevention (DGSMP) and the German Society for Epidemiology (DGEpi) first published "Good Practice in Secondary Data Analysis (GPS)" formulating a standard for conducting secondary data analyses. GPS is intended as a guide for planning and conducting analyses and can provide a basis for contracts between data owners. The domain of these guidelines does not only include data routinely gathered by statutory health insurance funds and further statutory social insurance funds, but all forms of secondary data. The 11 guidelines range from ethical principles and study planning through quality assurance measures and data preparation to data privacy, contractual conditions and responsible communication of analytical results. They are complemented by explanations and practical assistance in the form of recommendations. GPS targets all persons directing their attention to secondary data, their analysis and interpretation from a scientific point of view and by employing scientific methods. This includes data owners. Furthermore, GPS is suitable to assess scientific publications regarding their quality by authors, referees and readers. In 2008, the first version of GPS was evaluated and revised by members of AGENS and the Epidemiological Methods Working Group of DGEpi, DGSMP and GMDS including other epidemiological experts and had then been accredited as implementation regulations of Good Epidemiological Practice (GEP). Since 2012, this third version of GPS is on hand and available for downloading from the DGEpi website at no charge. Especially linguistic specifications have been integrated into the current revision; its internal consistency was increased. With regards to contents, further recommendations concerning the guideline on data privacy have been added. On the basis of future developments in science and data privacy, further revisions will follow.

Consideration of international generic distribution policies on patient outcomes in the United States and Germany.

Generic substitution of narrow therapeutic index drugs can have unintended consequences. Generic switching is often driven by cost incentives, regulations and supply, but may raise concerns about equal bioavailability, therapeutic equivalence and about possible confusion for the patient. Integrated systems of care with active management of patient behaviors, including adherence, may minimize the impact of switching. This article is intended to present policy drivers and potential consequences of generic switching and the role of pharmacist education in minimizing patient risk using warfarin and the pharmaceutical distribution systems of the United States and Germany as examples.

[Pain-specific diagnosis patterns in claims data - Identification by means of classification and regression trees (CART)]. / Schmerztypische Diagnosemuster in Routinedaten - Identifikation mittels Classification and Regression Trees (CART).

The identification of beneficiaries with persistent, recurrent or chronic pain in claims data by means of individual diagnoses or analgesic prescription is not sufficient and reliable. By using CLASSIFICATION AND REGRESSION TREES (CART) it was possible to identify specific diagnosis patterns for patients suffering from pain. Diagnosis patterns are considered as specific if they occur more frequently among beneficiaries with at least two opioid prescriptions within one year compared with beneficiaries who did not receive any analgesic therapy. Diagnosis and prescription data of 2006 were provided by the German sickness fund DAK. As a result, 65 diagnosis patterns occurred more frequently among beneficiaries treated with opioids than among the control group. These 65 patterns can be classified as follows: cancer-related pain (4), specific back pain/osteoporosis (8), spine-related pain (6), arthritis-related pain/rheumatoid arthritis (22), pain after traumatic fractures (5), pain in multimorbid, dependent patients (3), neuropathic pain (7), headache (5), non-specific back pain (5). The derived diagnosis patterns showed high predictive values (sensitivity: 78%, specificity: 66%) and are suitable for the identification of beneficiaries suffering from pain - the first step towards health services research in pain-based on claims data.

[Identification and grouping of pain patients according to claims data]. / Identifikation und Gruppierung von Schmerzpatienten anhand von Routinedaten einer Krankenkasse.

The ICD classification does not provide the opportunity to adequately identify pain patients. Therefore we developed an alternative method for the identification and classification of pain patients which is based on prescription and diagnoses data from the year 2006 of one nationwide sickness fund (DAK) and which is led by two main assumptions: 1. Beneficiaries without prescription of an analgetic drug but with a diagnosis pattern that is characteristic of patients who are treated with opioids are also likely to be pain patients. 2. Each combination of diagnosis groups can be traced back to one primary diagnosis out of a diagnosis group according to the patient classification system CCS (Clinical Classifications Software). The selection of this diagnosis group (CCS) allows for the allocation of the beneficiary to only one pain type. As a result we identified 65 combinations of CCS diagnosis groups--aggregated to nine "CCS pain types"--to which 77.1% of all patients with at least two opioid prescriptions can be allocated: 26.3% to pain due to arthrosis, 18.0% to pain due to intervertebral disc illnesses, 13.1% to other specific back pain, 6.7% to neuropathic pain, 4.5% to unspecific back pain, 4.2% to headache, 2.4% to pain after traumatic fractures, 1.3% to pain of multimorbid, high-maintenance patients, and 0.6% to cancer pain. Based on our method beneficiaries who have a high probability of suffering from moderate to strong pain can be identified and included in further claims data analyses of health care delivery and utilization pattern of pain-related disorders in Germany.

Economic burden of stroke: a systematic review on post-stroke care.

OBJECTIVES: Stroke is a leading cause for disability and morbidity associated with increased economic burden due to treatment and post-stroke care (PSC). The aim of our study is to provide information on resource consumption for PSC, to identify relevant cost drivers, and to discuss potential information gaps. METHODS: A systematic literature review on economic studies reporting PSC-associated data was performed in PubMed/MEDLINE, Scopus/Elsevier and Cochrane databases, Google Scholar and gray literature ranging from January 2000 to August 2016. Results for post-stroke interventions (treatment and care) were systematically extracted and summarized in evidence tables reporting study characteristics and economic outcomes. Economic results were converted to 2015 US Dollars, and the total cost of PSC per patient month (PM) was calculated. RESULTS: We included 42 studies. Overall PSC costs (inpatient/outpatient) were highest in the USA ($4850/PM) and lowest in Australia ($752/PM). Studies assessing only outpatient care reported the highest cost in the United Kingdom ($883/PM), and the lowest in Malaysia ($192/PM). Fifteen different segments of specific services utilization were described, in which rehabilitation and nursing care were identified as the major contributors. CONCLUSION: The highest PSC costs were observed in the USA, with rehabilitation services being the main cost driver. Due to diversity in reporting, it was not possible to conduct a detailed cost analysis addressing different segments of services. Further approaches should benefit from the advantages of administrative and claims data, focusing on inpatient/outpatient PSC cost and its predictors, assuring appropriate resource allocation.

The Impact of Generic Substitution on Health and Economic Outcomes: A Systematic Review.

BACKGROUND: Generic drugs are considered therapeutically equivalent to their original counterparts and lower in acquisition costs. However, the overall impact of generic substitution (GS) on global clinical and economic outcomes has not been conclusively evaluated. OBJECTIVE: To test whether (1) generics and original products yield the same health outcomes, and (2) generic therapies save economic resources versus original therapies. METHODS: We performed a systematic literature review in Medline, Embase, and the Cochrane Database of Systematic Reviews to identify original studies that examine clinical or economic outcomes of GS. After standardized data extraction, reported outcomes were categorized as supporting or rejecting the hypotheses. Each reported outcome was assessed and accounted for supporting and opposing GS. One publication could provide multiple outcome comparisons. RESULTS: We included 40 studies across ten therapeutic areas. Fourteen studies examined patients on de novo therapy; 24 studies investigated maintenance drug therapy, and two studies considered both settings. Overall, 119 outcome comparisons were examined. Of 97 clinical outcome comparisons, 67% reported no significant difference between generic drugs and their off-patent counterparts. Of 22 economic comparisons, 64% suggested that GS increased costs. Consequently, hypothesis (1) was supported but hypothesis (2) was not. We found no major differences among studies that investigated clinical outcomes with de novo or maintenance therapy. CONCLUSION: The review suggests that clinical effects are similar after GS. However, economic savings are not guaranteed. More systematic research comparing clinical and economic outcomes with or without GS is needed to inform policy on the use of generic substitution.

[Information the family physician gives his patients to take home. Utilization, contents and origin of printed information]. / Das gibt der Hausarzt seinen Patienten auf den Weg. Einsatzhäufigkeit, Inhalte und Herkunft schriftlicher Informationsmaterialien.

Providing the patient with information is one of the most important tasks of the family doctor, both for legal and medical reasons. Apart from verbal information, a wealth of printed material is also available. In a survey of 62 randomly selected general practices, the volume and nature of such information material was investigated, and its importance in practice established by means of an interview. 88.7% of the doctors interviewed stated that they handed out printed matter every day, and 96.8% considered the availability of such information to be an important aid in their daily work. Most of the printed information dealt with metabolic and cardiovascular disorders, and 93.5% of the material was supplied by the pharmaceutical industry. This means that other institutions which have a task to inform patients are virtually not represented in the doctor's office. Printed material providing health-promoting information, for example, on such topics as smoking, AIDS, exercise, healthy eating and suchlike, were lacking. The data point to an urgent need for additional information material in the doctor's office.

Excess health care costs of obesity in adults with diabetes mellitus: a claims data analysis.

BACKGROUND: Body weight management is a key factor in diabetes mellitus. However, both behavioral and pharmacological innovations to manage obesity may imply additional costs. In order to provide further insights into the role of obesity in diabetes-associated resource consumption, this study aims to estimate incremental costs of concomitant obesity in German adult patients (≥ 18 years) with different types of diabetes. METHODS: Adopting a third-party payer perspective, claims data from a German statutory sickness fund (N=1,094,496) were analyzed for costs of annual drug prescriptions and out- and inpatient care in adult beneficiaries with diabetes in 2004. Using diagnostic information, 37,570 beneficiaries with diabetes were identified. Concomitant obesity was assessed by ICD-10-codes (E66) in the claims data. Adjusting for sex, age, and micro- and macro-vascular complications, one generalized gamma regression model with the log link was performed for type 2 diabetes patients (N=24,562), type 1 diabetes patients (N=5,663), and an unclassified group (N=7,345), respectively. RESULTS: Overall, 33% of the patients with diabetes were identified as obese (type 2 diabetes: 34%, type 1 diabetes: 20%, unclassified: 38%). Affirming descriptive analyses, the generalized gamma regression models revealed that obesity is associated with significant increments in health care costs regardless of type of diabetes (type 2 diabetes: € 454, type 1 diabetes: € 812, unclassified: € 532). The interaction of obesity and macro-vascular complications was numerically stronger in type 1 than in type 2 diabetes but reached statistical significance only in type 2 diabetes (and the unclassified group). Moreover, concurrent macro- and micro-vascular complications were associated with higher incremental costs in all groups. CONCLUSIONS: Concomitant obesity is independently associated with incremental health care costs in adult patients with type 2 diabetes and, even more so, type 1 diabetes. Results are discussed with respect to the fact that in this sample, concurrent micro- and macro-vascular complications were more frequent in type 1 diabetes. At any rate, in light of these health care costs, obesity seems relevant in both types of diabetes. Due to claims data limitations, it was not possible to distinguish obesity classes based on body height and weight information. Further research should identify adiposity thresholds for increased resource consumption using both primary and secondary data.

[Pharmacoepidemiology. Analysis of drug prescription data]. / Pharmakoepidemiologie. Nutzung der Arzneimittelverordnungsdaten.

Due to an increase in pharmaceutical expenditure and growing demands on the observation and analysis of the pharmaceutical use in the overall population, pharmacoepidemiology has increasingly gained importance. In the foreground of this scientific discipline is the analysis of drug use in every day treatment; pharmaceutical prescription data are a suitable data source as they are part of the health care routine. In Germany prescription data are collected as process data of the Statutory Health Insurance (GKV) under naturalistic conditions and are for this reason not distorted by study designs nor artificial survey conditions. The contribution at hand states the most common data sources und gives an overview on the character of pharmaceutical prescription data, their information content and how they can be used for scientific research. The prevailing study designs which are applied in pharmacoepidemiologic surveys and the prevalent methodological approaches of the pharmaceutical appliance research on the basis of secondary GKV data are presented. Furthermore, an outlook is given on Pharmacoepidemiological Databases (PDb) that are already well-established in other countries and possibilities will be shown for monitoring drug utilization and turnover.

Epidemiology, treatment and costs of osteoporosis in Germany--the BoneEVA Study.

INTRODUCTION: In Germany, accurate data on the prevalence and treatment of osteoporosis, as well as the cost of this illness, are not available. The aim of this study is to give a valid estimation of these items for the year 2003. METHODS: Routine data from a German sickness fund covering 1.5 million beneficiaries and billing data for outpatient visits were used to obtain estimates of prevalence for osteoporosis. Claims data for patients with osteoporosis (M80, M81) or an osteoporosis-related fracture diagnosis (S22, S32, S42, S52, S72, S82) or treatment with anti-osteoporosis drugs were examined. Costs were calculated from the perspective of the German health insurance system and the German nursing care insurance system, respectively. Only direct costs of osteoporosis were considered. RESULTS: In 2003, 7.8 million Germans (6.5 million women) were affected by osteoporosis. Of them, 4.3% experienced at least one clinical fracture. Only 21.7% were treated with an anti-osteoporosis drug. The total direct costs attributable to osteoporosis amounted to euros 5.4 billion. CONCLUSION: This study confirms that osteoporosis is underdiagnosed, undertreated and imposes a considerable economic burden on the health system in Germany. Effective strategies for the prevention and management of this disease are needed.

Benzodiazepine substitution in medical practice. Analysis of pharmacoepidemiologic data based on expert interviews.

Pharmacoepidemiologic data have shown a consistent reduction in benzodiazepine (BZ) prescriptions over the past decade. The question remains as to whether BZs are simply put aside or whether other medications are used as substitutes. Expert interviews, at which a stratified sample of 114 psychiatrists, internists, and general practitioners were presented with case studies, were conducted to learn about the therapeutic alternatives in the field of benzodiazepine-related indications administered in daily practice. These results were used to analyze trends in pharmacoepidemiologic prescription data for all patients under the general health insurance plan from 1981 to 1988. The experts identified distinct alternatives to BZs in different clinical situations, including neuroleptics, antidepressants, phototherapeutics, and analgesics. When these findings were transferred to pharmacoepidemiologic data, results revealed an increase in the prescription of alternative medications that apparently compensated for reduced BZ use. Overall there was no change (or, rather, no increase) in the total of psychotropic prescriptions during the period of reduced BZ prescriptions. Our findings indicate that reduction in benzodiazepines prescription is associated with substitution by various other psychotropic drugs. This has positive as well as negative consequences, and there must be discussed in detail before sound recommendations can be given as to which type of drug. BZs included, is the best choice in which type of illness.

Benzodiazepines in medical teaching. A survey of university departments of pharmacology, psychiatry, and internal medicine in Germany.

Benzodiazepines are one of the most frequently prescribed classes of medications and by far the biggest class of psychotropic drugs. For years they have been under critical discussion both professionally and in public. The questions arise as to what view is held by university teachers on this topic and how much attention is given to benzodiazepines and benzodiazepine-related problems. Have there been changes in medical teaching in recent years? All the West German departments of psychiatry, internal medicine, and pharmacology were asked by means of a mailed questionnaire to report on their teaching practice. 84.8% of the questionnaires were returned. Results show that mainly psychiatrists and pharmacologists place special emphasis on benzodiazepines in medical teaching. Additionally, in recent years they have devoted more time to this drug class and to related risks in particular. In particular, psychiatrists deal more with benzodiazepines in postgraduate training and the further training of physicians as well as in their scientific research. When the responses of university teachers in psychiatry, internal medicine and pharmacology are summed up, they show that chiefly psychiatrists have responded in their teaching to the current discussion. This contrasts particularly with internists. This has to be discussed in the light of the fact that, in terms of quantity, minor tranquilizers are predominantly prescribed by nonpsychiatrists.

Specialty training and the personal use of benzodiazepines by physicians affect their proneness to prescribe tranquilizers.

The decision on how to treat a patient does not depend on clinical matters or illness characteristics alone, but also on patient, physician and setting variables such as personality, training, or reimbursement. No research has yet been carried out to answer the question whether personal experience with medications also influences prescribing behavior. In this study, 124 physicians stratified according to specialty (neuropsychiatrists vs. general practitioners), type of institution (private practice vs. hospital), years of professional experience (young vs. old), and region (rural vs. urban) participated in a structured interview to evaluate their proneness to prescribe benzodiazepines for sleep disorders as well as their personal experience in taking benzodiazepines for their own sleep problems. Both specialty and personal experience were significantly related to proneness to prescribe. Other variables tested (region, institution, age, gender) did not help to explain the variance in benzodiazepine prescribing practice. Thus physician variables and, importantly, their own personal experience in taking the medication significantly influence treatment choice. Rational medical decision making and treatment guidelines must therefore take into account medical knowledge as well as knowledge of personal treatment preferences and professional biases.

Pathways to care and psychological problems of general practice patients in a "gate keeper" and an "open access" health care system: a comparison of Germany and the Netherlands.

BACKGROUND: The comparison of different health care systems is one way to give empirical evidence to health care reform and policy. The differences between health care systems in which general practitioners serve as gate keepers in comparison to systems in which patients are free to contact every physician and specialist they like are a question of high interest. METHOD: This study compares the Netherlands and Germany, two countries with very similar political, social, and health system structures, but different types of access to the health care system. While Germany offers unconstrained access to specialist ambulatory care, the Netherlands restricts health care utilization by giving primary care a 'gate keeper' function not allowing patients direct access to specialist care. Data from the WHO international collaborative study on psychological problems in general health care (Ustün and Sartorius 1995) were analysed with respect to pathways to care, treatment, and health status. In an initial cross-sectional assessment, in 3-month and 12-month follow-ups, contacts to physicians or hospital admission have also been monitored. RESULTS: There were only marginal differences between the Dutch and the German sample in the sociodemographic characteristics as well as in the diagnostic status with respect to mental disorders. In the Netherlands, 95.5%, and in Germany, 68.8% of the patients presented their 'reason for visit' for the first time to any physician at this index contact with a general practitioner. During the following 3 months, 24% of the Dutch patients, but 60.2% of the German patients, additionally contacted other physicians ( P < 0.001). At 12 months, this rate was 62.9% vs. 78.6% ( P < 0.001). During the 12-month follow-up period, there were 15.7 0/00 hospital admissions in Germany vs. 25.4 0/00 in the Netherlands ( P < 0.005) [corrected]. CONCLUSIONS: Family physicians in a gate keeper system reduce the number of contacts to other physicians and the intensity of treatment, while at the same time the rate of hospital admissions is increased.

What happens to patients after the end of a clinical trial? Systematic follow-up observational study of an open moclobemide trial in major depression.

Controlled clinical drug trials typically last several weeks. At the end of this fixed time period approximately two-thirds of the patients in trials with antidepressant drugs are classified as responders, i.e. the initial severity score is reduced by about fifty percent. In a clinical perspective this means that many patients, even among the responders, are still quite ill and require further treatment. Nevertheless, to date there do not appear to have been any studies on treatment and course of illness after the termination of controlled clinical trials. In a "naturalistic" follow-up surveillance study of 202 patients who had taken part in a controlled trial with the antidepressant moclobemide for six weeks, treatment and clinical status were monitored through questionnaires sent out to their treating physicians (n = 78) at two assessment points four weeks and six months later. Results showed that sixty percent of patients continued to receive the study drug moclobemide after the completion of the clinical trial. Twenty percent were switched to other antidepressants and twelve percent received no further psychotropic drugs. In the course of six months a considerable variation in treatment modes could be observed. Patients under ongoing active treatment made considerable additional progress. A correlation was found between insufficient response and repeated switching of medication. Patients who had been taken off antidepressant medication because of early good response did not experience early relapse after the end of treatment.

Early termination of antidepressant drug treatment.

Guidelines for antidepressant drug therapy recommend continuing treatment for several months, even after recovery. Premature termination of medication is viewed as a significant reason for chronicity in depressive disorders. To study the scope and type of this problem, routine treatment cases must be observed in sufficient numbers. In four subsequent drug utilization observation studies, the time course and reasons for early termination (i.e., during the first weeks of fluoxetine treatment for depression) were investigated in (1) 15,601 patients of 4,696 general practitioners; (2) 2,401 patients of all ages treated by 479 neuropsychiatrists from 1990 to 1991; (3) 757 patients of all ages treated by 213 neuropsychiatrists from 1992 to 1993; and (4) in 977 geriatric subjects treated by 216 neuropsychiatrists. In study 1, treatment was terminated during the first 10 weeks in 32.9% of cases; in study 2, in 48.0%; in study C, in 31.2%; and in study D, in 31.0%. When treatment was discontinued because of adverse events, the median of treatment duration was approximately 15 days; in cases of clinical deterioration, 20 days; in cases of poor response, 40 days; and when the reason was good response, the median of treatment duration was approximately 43 days. This is the largest field study on early termination of antidepressant treatment available to date. The consistency of results, which were collected from a large number of patients of different ages and from specialists as well as general practitioners, speaks for the validity and stability of the observed phenomena. The different forms of early treatment termination are discussed and addressed separately because they each have special risks and are differently associated with treatment failure.