VersKiK: Study protocol of an observational registry-based study on the current state of follow-up care and adherence to follow-up guidelines after cancer in childhood or adolescence

BACKGROUND

This article describes the study design of the quantitative part of the VersKiK study, The primary objectives of this study are to examine the occurrence of late effects in survivors of childhood or adolescent cancer (module 1), investigate health-related vulnerabilities and medical service utilization within this survivor group (modules 1 and 3), and assess the alignment between documented follow-up care for cardiological and audiological late effects with guideline recommendations, along with evaluating the extent of adherence among paediatric cancer survivors (module 3).

METHODS

This is a non-interventional retrospective observational cohort study. It is based on stochastically linked insurance claims data from approximately 150,000 statutory insured persons with information concerning around 25,000-30,000 cancer survivors recorded in the German Childhood Cancer Register (GCCR). To explore adherence to selected follow-up guidelines, intention to treat treatment data from clinical study groups for particular diagnostic entities will be additionally included.

DISCUSSION

The growing group of survivors after cancer in childhood and adolescence is representing a special population with an increasing demand for life-long healthcare services through relative high probability of late effects. Currently, there is a limited evidence in Germany on utilization of corresponding medical services and adherence to follow-up guidelines. With this study design, we are aiming to address these gaps and, consequently, suggest improvements to existing follow-up guidelines and follow-up care provision in Germany.

Epidemiology of cardiac amyloidosis and burden on health care system in Germany: a retrospective analysis from 2009 to 2018

Aims

Improved imaging modalities contributed to increasing disease awareness of cardiac amyloidosis. Contemporary data on frequency trends and impact on health system are lacking.

Methods and results

In a retrospective study using health claims data of the largest German statutory health insurance patients with a diagnostic code of amyloidosis and concomitant heart failure or cardiomyopathy between 2009 and 2018 were identified. Temporal trends in frequency, patient characteristics, all-cause mortality and measures of health care burden were examined.

8,279 patients were identified of whom 5,618 were incident without diagnosis of amyloidosis within the previous year. Prevalence increased from 15.5 to 47.6 per 100,000 person-years, and incidence increased from 4.8 to 11.6 per 100,000 person-years, with a continuous steepening in the slope of incidence trend. Age and male gender significantly increased whereas prevalence of myeloma and nephrotic syndrome significantly decreased over time. Median (IQR) survival time after first diagnosis was 2.5 years (0.5 to 6 years), with a 9% (95% CI 2–15%, p=0.008) reduced risk of death in the second compared to the first five years of observation. In the first year after diagnosis mean total health care costs were 21,955 € (median 9,873 €, IQR 3,922 to 24,714 €) per person.

Conclusion

The rise in patients with cardiac amyloidosis has continuously accelerated in the last decade which, based on patient characteristics, is mainly driven by underlying wildtype transthyretin amyloidosis. Considering the adverse outcome and high health care burden further effort should be put on early detection of the disease to implement treatment.

Funding Acknowledgement

Type of funding sources: None.

Long-term care, care needs and wellbeing of individuals after cancer in childhood or adolescence (VersKiK): study protocol of a large scale multi-methods non-interventional study

BACKGROUND

It has been shown previously that a relevant proportion of childhood cancer survivors suffers from late effects, which are often directly related to the cancer itself or its therapy, resulting in particular follow-up needs, additionally burdening healthcare systems. Being diagnosed with cancer at a vulnerable stage of development, this group of cancer survivors is at comparatively higher risk of relapse or subsequent cancer. Although national and international follow-up guidelines based on treatment modalities have been developed, their implementation seems to leave room for improvement. Additionally, they lack a sufficient consideration of the survivors' psychosocial needs, affecting their adherence to them. The aim of the VersKiK study is to provide representative information on late effects in childhood and adolescence cancer survivors in Germany. The main research objectives are: (1) to describe the state of follow-up care among survivors after a cancer diagnosis in childhood or adolescence; (2) to quantify the occurrence of late effects among this group of survivors; (3) to examine the adherence to selected audiological and cardiological follow-up guidelines and to identify factors affecting it; (4) to explore actual follow-up needs of paediatric cancer survivors; (5) to review selected follow-up guidelines with the aim to improve and expand them.

METHODS

VersKiK is designed as a mixed-methods non-interventional study. We will use claims data from statutory health insurance companies in combination with individually linked population-based registry data from the German Childhood Cancer Registry (GCCR). This data base will permit us to quantify diagnoses and procedures in comparison to the general population as well as the adherence to existing follow-up guidelines. Additional information will be obtained through interviews with childhood and adolescence cancer survivors and their informal caregivers, as well as in focus groups with healthcare professionals.

DISCUSSION

The present study aims to research the actual needs of individuals after cancer diagnosis and treatment in childhood or adolescence - physical, psychological and organisational - in order to improve existing follow-up guidelines. These improvements might further positively affect not only actual care provided to paediatric cancer survivors, but also benefit healthcare systems in general while decreasing consequent medical visits in this group of patients.

TRIAL REGISTRATION

Registered at German Clinical Trial Register (ID: DRKS00025960 and DRKS00026092).

[Rehabilitation in geriatric patients after ischemic stroke--a comparison of 2 organisational systems in Germany using claims data of a statutory health insurance fund]

Due to historical aspects in some federal states in Germany rehabilitation of geriatric patients is organized in geriatric departments in hospitals (§ 109 SGB V). In other states rehabilitation of these patients is mainly realized in geriatric rehabilitation facilities outside hospital care after approval by the health insurance (§ 111 SGB V). Thus, it is of interest to compare both types of health care with respect to differences in population characteristics, resource utilization and outcome parameters (i.e., excess costs, rehospitalization, fracture risk and mortality) using a common geriatric indication, the ischemic stroke, as an example.

METHODS

Claims data of the AOK (Local Health Care Fund) from seven federal states in Germany were used. Insured persons with a documented hospital stay with discharge diagnosis cerebral infarction/stroke (ICD-10 I63, I64, below denoted by "ischemic stroke") in 2007 (N=39,887) were included and allocated to the respective form of rehabilitative health care via the OPS (German procedure classification for inpatient procedures) procedure 8-550 (§ 109, N=1,272) or via admission to a geriatric rehabilitation unit within 1 month after hospital discharge (§ 111, N=2,200). All direct costs were ascertained and presented with and without costs of long-term care. Excess costs were calculated as the difference of costs between the first year after insult and the costs in the previous year. Excess costs in the 2 types of care were compared using multivariate quantile regression analysis. Risk of hospitalization (due to ischemic stroke or fracture) and risk of death in a 1-year follow-up was analysed using multivariate cox regression.

RESULTS

Insured members treated according to health care type § 109 were somewhat older (mean: 81 vs. 80 years of age), more frequently female (72 vs. 67%), more often receiving long-term care (27 vs. 19%) and had more often documented sequelae after insult (>=4 diseases 39 vs. 28%). No significant differences in excess costs between both types of care were observed (quantile regression: 25%-percentile-comparison: p=0.49 and 0.11; median-comparison: p=0.99 and 0.13; 75%-percentile-comparison: p=0.13 and 0.30, with and without costs of long-term care, respectively). Moreover, no significant differences were observed related to the outcomes 'rehospitalization due to ischemic stroke' (hazard ratio - HR [95% confidence interval - CI])=1.12 [0.85-1.48], p=0.43) and death (HR [95% CI]=1.03 [0.88-1.20], p=0.75) in the multivariate model (reference: health care type § 111). Insured members in health care type § 109 had a significant lower risk of rehospitalization due to fracture (HR [95% CI]=0.61 [0.40-0.93], p=0.02).

CONCLUSION

According to health care type § 109 and § 111, geriatric patients differ in certain characteristics such as gender, statutory care and documented sequelae after insult. Except for the outcome 'fracture', no significant differences between both types of care have been observed in the selected outcomes. Primary studies with more differentiated data collection may focus on specific treatment and on aims and achievements of rehabilitation.

Colonoscopy and fecal occult blood test use in Germany: results from a large insurance-based cohort

BACKGROUND AND STUDY AIMS

Colonoscopy and guaiac-based fecal occult blood tests (FOBT) are recommended and offered for colorectal cancer (CRC) screening in Germany. We aimed to explore their utilization in a large insurance-based cohort.

PATIENTS AND METHODS

Claims data from between 2000 and 2008 were collected for 170493 individuals who were insured by a large health insurance plan in the federal state of Hesse, Germany. The percentages of individuals who had recently utilized CRC screening-related procedures were calculated. Additionally, multiple test use and identification of CRC screening providers were ascertained.

RESULTS

Following the inception of the current CRC screening program in 2002, colonoscopy utilization rates varied only slightly and FOBT use decreased in individuals aged ≥50 years. At the end of 2008, the age-standardized percentages of individuals who had undergone colonoscopy within ≤10 years were 23% for men and 26% for women. The proportions of individuals who had used FOBT within ≤1 year were 14% for men and 22% for women. Patient education had been utilized by 38% of eligible persons and was increasingly followed by screening colonoscopy. For women, practices that specialized in gynecology were the main providers of FOBT (93%) and patient education (61%).

CONCLUSIONS

This study provides new insights into the inter-related utilization of colonoscopy, FOBT, and patient education in Germany, and may be particularly informative for the design of strategies to increase CRC screening uptake. It indicates that sex differences in CRC screening test use could result to a large extent from general visits to different types of specialist physicians involved in the CRC screening process.

[Differences in antibiotic prescribing between general practitioners and pediatricians in Hesse, Germany]

BACKGROUND

A careful consideration of antibiotic prescription is necessary due to emerging antibiotic resistance. Differences in prescription rates between physician groups have only rarely been examined. We compared the prescription of antibiotics for systemic use in children between pediatricians and general practitioners (GPs).

METHOD

Data source was the statutory health insurance sample AOK Hesse/KV Hesse. Overall, 47,033 insured children in 2006 between 2 and 17 years of age were included in the analysis. Odds ratios (OR) and 95% confidence intervals (CI) for an antibiotic prescription were calculated by logistic regression adjusted for confounders, comparing pediatricians with GPs. Diagnosis-specific models were constructed.

RESULTS

GPs and pediatricians prescribed 79.3% of all antibiotics in 2006 (40.6% and 38.7%, respectively). Significantly decreased ORs for an antibiotic prescription were observed for pediatricians, with OR (95% CI) of 0.46 (0.43-0.49), 0.40 (0.30-0.54) and 0.40 (0.30-0.53) for respiratory tract infections, urinary tract infections and nonsuppurative otitis media, respectively. No significant associations were observed assessing scarlet fever, pneumonia and suppurative otitis media, respectively.

CONCLUSION

In this retrospective analysis, pediatricians were associated with a lower antibiotic prescription rate regarding diagnoses where antibiotic therapy is not clearly indicated. However, in diagnoses where antibiotic therapy is clearly indicated, no differences in prescription rates between pediatricians and GPs were observed. Further studies are warranted to gain insight into different treatment strategies between different groups of medical practitioners.

[Data protection and methodological aspects in compiling a routine database from statutory health insurance data for research purposes]

Personally identifiable routine data generated by the SHI (statutory health insurance) offer inexpensive and large amounts of data gathered over long periods of observation for use in numerous fields of application including health services research and epidemiology of health care. As a source of medical health information, these data are subject to particular EU data protection directives according to which they can only be used under certain conditions and following careful consideration of the various interests involved. These interests include the protection of personal privacy, on the one hand, and the freedom of research, on the other. As personally identifiable data, these data are fully subject to general and specific data privacy regulations, such as the consideration of intended use; the specification of forms of data processing, duration of use, and group of users; and the development of a data protection concept. If primary data are additionally collected, the patient is to be fully informed about the intended contents of analysis and the use of his/her data in order that informed consent can be provided. Methodological standards such as the verification of completeness and plausibility are also to be met when compiling an insuree database.

[Impact of demographic changes on health care expenditure in Germany: an analysis considering the expenditures of decedents]

The purpose of this study was to evaluate the impact of demographic changes on future health care expenditure of the German social health insurances considering the expenditures of survivors and decedents by age. The study analysed data from 269,646 members up to the age of 99 years of the AOK - one of Germany's largest social health insurers - in the State of Hesse in 2000/2001. In order to determine future health care expenditures, per-capita expenditures by age for outpatient, inpatient, rehabilitation, and nursing services of survivors and decedents (death within the next 12 months) were multiplied by the estimated number of survivors and decedents by age in Germany in 2020, 2035 und 2050. Expenditures for all ages were summed together. The paper shows that demographic changes until 2050 will lead to an increase of health care expenditures by 20% in total or less than 1% annually. Considering the future re-duction in workforce, demographic changes until 2050 will result in an estimated increase in health care expenditures per employee by about 57% (undifferentiated model). Considering the cost of survivors and decedents separately, this increase will amount to 50%. Hence, undifferentiated models overestimate the impact of demographic changes by about 10%.

[Persons requiring long-term care: Recommendation and utilization of rehabilitative therapies]

In Germany all members of the statutory health insurance are also compulsory members of the nursing care insurance which financially supports nursing care at home and in nursing homes. Benefits are provided dependent on a standardized medical examination performed at the home of the applicants by trained physicians or nurses of the health insurances' medical service (MDK). The benefits are granted to those persons who are limited in the performance of activities of daily living due to physical, cognitive or mental disorders. In the German Code of Social Law regulating the long-term care system, one important entitlement is "rehabilitation before long-term care". It aims at maintaining life of disabled persons in their familiar surroundings as long as possible. Up to now in the nursing care insurance this aim has been implemented insufficiently. The objective of this study was therefore to investigate the frequency of suggested ambulatory rehabilitation, which is recommended by the MDK at the examinations, and the following prescription of ambulatory rehabilitation, issued by an office-based physician. The database included the medical examinations of the years 2001 to 2002 linked to a random sample of persons insured in a statutory sickness fund in the German federal state of Hesse. Nursing care-related information including recommended rehabilitation therapy was derived from the medical examinations. Insurance data provided information on prescriptions. Data of 7,840 persons could be analyzed. Rehabilitative therapy was recommended in 15% of the nursing care applicants, mostly physiotherapy (ca. 90%). Persons receiving the recommendation were mainly male, they were younger, were more disabled, had less often a diagnosis of the ICD-10 chapter "symptoms and abnormal clinical and laboratory findings" or "mental and behavioral disorders", live less often alone and were more restricted in their motion compared to persons without recommendation for rehabilitative therapy. Only 55.3% of the persons with a recommendation for physiotherapy and only one-third with a recommendation for occupational therapy or logopedics received a prescription for the respective therapy in the following three months after examination. The age younger than 80 years, professional care, no home care, and-as the strongest item-preceding therapy were positively related to receiving a prescription. These findings show that disabled persons have the capability for prevention and rehabilitation, but it is used in only half of them. One possible reason might be the financing of the remedies, which burdens the budget of the health insurance, where instead the nursing care insurance benefits from the possible delay of the utilization of nursing care.

Cost of illness of malignant lymphoma in Germany

Cancer causes a high economic burden. The purpose of this study is to determine and compare the direct, indirect and societal costs of illness for Hodgkin's Disease (HD), Non-Hodgkin's Lymphoma (NHL), Plasmocytoma and Chronic Lymphatic Lymphoma (CLL). We used a database of 1.9 million individuals enrolled in a statutory sickness fund in Germany to identify 4,172 patients treated for malignant lymphoma in 2000. Direct, indirect and societal costs were calculated using a case-control design and the human capital approach. Direct cost (in Euro) for patients with HD was 3604, for NHL patients 6,149, for Plasmocytoma 8,400, and for CLL patients 3,226. Total indirect cost for HD was 69 million, for NHL patients 404 million, for Plasmocytoma 144 million, and for CLL patients 52 million. Totalling 1.7 billion Euro in economic cost for Germany in 2000, with 44,000 productive years lost, malignant lymphomas are a relatively costly disease group. As life expectancy increases, costs for malignant lymphoma are likely to rise due to the high prevalence among the elderly. Further research employing disaggregated, incidence-based cost is needed.

The cost burden of diabetes mellitus: the evidence from Germany--the CoDiM study

AIMS/HYPOTHESIS

The aim of this study was to identify the health care costs of diabetic patients in Germany in 2001, focusing on the influence of age, sex, and type of treatment.

SUBJECTS AND METHODS

Annual direct costs of medical care and indirect costs of inability to work and early retirement in diabetic subjects were compared with costs of age- and sex-matched non-diabetic control subjects. The analysis was based on routine health care data from a random sample (18.75%) taken from a database of 1.9 million insured persons. Incremental differences in medical and national expenditure between subjects with and without diabetes were calculated.

RESULTS

Annual direct mean costs per diabetic patient were 5,262 Euro, and indirect costs were 5,019 Euro. In the control group, mean direct and indirect costs were 2,755 Euro and 3,691 Euro, respectively. Analysis of cost components revealed that the high costs associated with the care of diabetic patients could be largely attributed to inpatient care and overall medication costs. Hypoglycaemic drugs amounted to only one-quarter of the medication costs. The total health care costs were correlated with the type of treatment. Direct excess costs increased with increasing age in insulin-treated patients, but were unaffected by age in patients receiving other types of treatment.

CONCLUSIONS/INTERPRETATION

The Costs of Diabetes Mellitus (CoDiM) study is the first comprehensive study to provide estimates of costs associated with diabetes care in Germany. Direct costs of diabetic patients account for 14.2% of total health care costs, which includes the proportion that specifically accounts for diabetes-related costs (6.8%).

Verläufe von Pflegebedürftigkeit in Hessen in den Jahren 1999 bis 2002. Ergebnisse einer Längsschnittstudie

AIM OF STUDY

This paper looks at pathways through care for care-dependent persons (as defined by the German statutory long-term care insurance) through longitudinal and cross-sectional analysis of types of benefits and benefit levels over a period of 4 years.

METHODS

The prospective cohort study with back dated recording consists of monthly data on the type and level of benefit for each care-dependent person. Cohorts of care-dependent people (n = 6.928) are followed over time to record the benefit type and level at program entry and during enrollment. Each change of benefit type or level, including exit from the program such as in the case of death is noted for a time period of 48 months.

RESULTS

The 4-years longitudinal study (1999 through 2002) proves important differences between courses of long-term care in relation to benefit types. Recipients of the cash benefit remain for a longer period of time in that benefit type (28.1 %), have lower transition rates to institutional care (6.9 %) and lower mortality rates (57.3 %) in comparison to recipients of other benefit types over the course of 4 years. Only 8 % respectively 10.7 % of recipients of the combined or the service benefit have the same type of benefit after 48 months. Transition rates to institutional care for professionally cared persons are 10.1 % (combined benefit) respectively 11.8 % (service benefit). Mortality rates of other than cash benefit recipients are about 70 %.

CONCLUSIONS

Further research should consider risk factors for entry into institutional care and mortality rates as well as questions regarding the prevention of care dependency and effectiveness of rehabilitation. Pathways through care should be analysed by focusing social variables like sex and chronic diseases of older people. Secondary analysis of process data from German statutory long-term care insurance program is a powerful tool for the study pathways through of long-term care.

Versichertenstichprobe AOK Hessen/KV Hessen - Konzeption und Umsetzung einer personenbezogenen Datenbasis aus der Gesetzlichen Krankenversicherung

Statutory health insurance data are being increasingly used for secondary data research. Longitudinal data can be prepared for research in health care, epidemiology or demand planning, in particular through the person-related nature of the data which is a precondition for the creation of inter-sector and inter-period data sets. This application possibility was introduced in a method study "person-related sampling of statutory health insurance data" and is now translated into practice on a larger scale for the first time in the regional sample "Versichertenstichprobe AOK Hessen/KV Hessen". For the collection and use of these data, model procedures were designed which take account of organisational (data access, contractual agreement, advisory board), technical (sampling, collection and storage of data) and confidentiality (data protection concept, pseudonymisation) aspects. The insured person-related sample may thus serve as a basis for the data pool planned for the national health system (Social Security Regulation 303 a-SGB V).

GPS - Gute Praxis Sekundärdatenanalyse

The scientific use of secondary data, especially of claims data from health insurance funds, has continuously increased in the last years. Therefore the Working Group "Collection and Use of Secondary Data" (AGENS) of the German Society of Social Medicine and Prevention (DGSMP) took the initiative to define quality standards for secondary data analysis. Starting with a review of the Good Epidemiologic Practice (GEP) AGENS adapted the GEP to the specific requirements of secondary data analysis by a multi-stage consensus process. The guideline Good Practice Secondary Date Analysis (GPS) was adopted on January 15 (th), 2005. GPS consists of 10 guidelines which are divided in explaining comments and recommendations. The GPS are targeted to set up standards for secondary data analysis, and they may also be used as a foundation of contracts between data owners and scientists. They are addressed to scientists from health services research and social medicine. AGENS commits itself to revise GPS continuously.

[Prospective comparative observational study on the antibiotic treatment of secondary peritonitis in Germany -- efficacy and cost analysis]

AIMS

Cephalosporins and broad-spectrum penicillins in combination with metronidazole are suitable for treatment of secondary peritonitis. The aim of this study was to compare the clinical and bacteriological efficacy, length of hospital stay (LOS), treatment costs of ceftriaxone (CRO) in combination with metronidazole vs standard regimens (SR) 1 and 2.

METHODS

Patient data were subjected to matched-pairs analysis according to four different categories of the Mannheim peritonitis index. RESULTS. From January 1998 to March 2000, a total of 365 patients from 59 surgical wards in German hospitals were included. Clinical efficacy was 90.2% vs 70.4% ( P=0.004) for CRO/SR 1 and 78.3% vs 82.6% for CRO/SR 2. Bacteriological efficacy was comparable. Antibiotic treatment costs were 593/539 Euros for CRO/SR 1 and 466/750 Euros for CRO/SR 2, i.e., costs for CRO were 37.9% lower than with SR 2.

CONCLUSION

Based on clinical, bacteriological, and pharmacoeconomic results, ceftriaxone in combination with metronidazole can be regarded as a first-line antimicrobial treatment of secondary peritonitis.

Development of indicators for assessing the quality of prescribing of lipid-lowering drugs: data from the pharmacotherapeutic quality circles in Hesse, Germany

OBJECTIVE

To develop indicators based on prescription analysis in order to assess adherence using guidelines and to monitor prescribing behavior. SETTING AND MATERIAL: Eleven pharmacotherapeutic circles (PTC) of the Association of Statutory Health Insurance Physicians (KV Hesse; 10 PTCs including 155 GPs as participants, mostly high prescribers; one circle with 11 moderators trained in pharmacology). These provided a total of 183,997 drug prescriptions involving 54,970 patients (prescriptions reimbursed by the substitute fund--Ersatzkassen--II. quarter 2000); claims form from 151 GPs. On average, 5.1% of the patients with prescriptions received lipid-lowering drugs.

METHOD

Development and application of indicators based on the guideline for the treatment of hypercholesterolemia developed by a GP's guideline group of the quality circles in Hesse (Hausärztliche Leitliniengruppe Hessen). The ratio of prescribing for primary and secondary prevention was chosen as a top indicator for measuring adherence to the guideline. Prescribing for secondary prevention was assessed by a set of special diagnoses. The second indicator relates to patient groups (here: older than 75 years) where the benefit of prescribing is not clear. Further indicators measure the prescribing of first choice drugs, avoidance of risk combinations and costconscious prescribing. The indicators are presented in the prescription analysis and will be discussed during the circle meeting.

RESULTS

On average, the moderators prescribed 34% of the lipid-lowering drugs for primary prevention, the GP circle participants 36.7%. On average, moderators and GP participants prescribed lipid-lowering drugs for 4.9% and 7.5% of patients older than 75 years, respectively (6% and 22% in primary prevention). Approximately, 28% of all lipid-lowering drugs issued by the participants involved simvastatin and pravastatin as first-choice drugs compared to 36.5% in the case of the moderators. The prescribing of statins with fibrates or macrolides in combination is seldom. Cost-conscious prescribing can be assessed for each GP by the percentage of generic prescribing and the number of different brands per active agent. Both, moderators and participants used generics when prescribing fibrates, bile acid sequestrants or other types of lipid-lowering drugs (moderators 53.8% and GP participants 78.5%). Three months is too short a period for assessing compliance of lipid-lowering drug prescribing.

CONCLUSION

It is possible to derive indicators from the guidelines and to integrate them into prescription analysis. The indicators reveal prescribing problems. The evaluation of PTCs in 2002 will involve the use of indicators as an instrument to assess the success in the implementation of guidelines.

[Confidentiality within the scope of secondary data research--approaches to a solution of the problem of data concentration]

In order to implement general data protection requirements and internationally recognised ethical requirements, research with personal health and social data demands a specific framework for the secure handling of confidential data. In the process of transferring data from the health service providers to the place where they are analysed, an important role is played by a so-called trust centre, responsible for pseudonymisation of personal and institutional identifiers. An undesirable concentration of data in the trust centre can be avoided by early separation of data in the data transfer institution: the trust centre receives only the identifier to be pseudonymised, while the health provision data are sent direct to the analysing institution, where they can be matched with the pseudonyms from the trust centre, with the help of a unique case number. The possibility of reidentification, which exists mainly in large (pseudonymised) data sets, can be reduced by use of an appropriate pseudonymisation process (e.g. insuree-based pseudonymisation by health service providers for sampling of insurees). The measures described here are suitable for protecting confidentiality and for further improving data security in the handling of confidential personal and institutional data.

Pharmacotherapeutic circles. Results of an 18-month peer-review prescribing-improvement programme for general practitioners

OBJECTIVE

To assess the effectiveness of the pharmacotherapeutic circle (PTC), a general practitioner (GP) prescribing-improvement programme to enhance prescribing quality and reduce drug costs.

DESIGN

Combined pre- and post-intervention time-series design using an internal comparison of subgroups and an external comparative control.

SETTING

Small discussion groups meeting 8 times over 18 months.

PARTICIPANTS

79 GPs exceeding the mean drug costs/patient of all Hessian physicians by > or = 40%; 10 moderators.

INTERVENTIONS

Peer-review feedback of prescription patterns based on guidelines targeting 3 suboptimal prescribing areas: drug prescriptions lacking evidence-based efficacy (target A); presumptive prescribing habits (target B); and underprescribing of new, effective therapies (target C).

MAIN OUTCOME MEASURES AND RESULTS

Significant decreases in prescription rates for target A drugs were recorded for varicose vein medications (p = 0.006), peripheral vasodilators (p = 0.0001) and topical antirheumatics (p = 0.0145), but not for prokinetics/enzymes/digestives. Prescribing of target B drugs such as benzodiazepines and nonsteroidal anti-inflammatory drugs declined markedly (p = 0.0019 and 0.0014, respectively). Target C drug prescriptions such as for opioids and proton pump inhibitors were not significantly increased. Highly significant reductions in prescription costs were observed for target A and B drugs, irrespective of whether GPs were stratified into high, medium or low prescribers. When mean prescribing costs for PTC participants were compared with those of a control group comprising 8000 GPs over a 21-month period, PTC GPs decreased their costs by 2%, whereas drug costs for all Hessian physicians rose by 10%.

CONCLUSIONS

PTCs appear to be an effective method to optimise the quality of drug prescribing and reduce drug costs.

Markers to analyse the prescribing of non-steroidal anti-inflammatory drugs in ambulatory care. A guide to pursuing rational and safe prescribing

OBJECTIVE

The aim of the study is to construct quality markers for rational prescribing of non-steroidal anti-inflammatory drugs (NSAIDs MO1A, MO1B) in such a manner that they meet three requirements: suitability to be assessed by prescription analysis, application in feedback strategies and contribution to the task of internal quality assurance.

METHOD

Eight different markers for validating NSAID prescribing were developed according to pharmacological literature. The prescribing of 99 participants (high prescribers) and 15 coordinators of eight pharmacotherapy circles (second quarter 1996) served as the database. To test the validity of the markers in terms of rationality, the NSAID prescribing of 15 randomly selected participants of these circles, whose participants were not trained in the analysis of their prescribing, was compared with the prescribing of the 15 coordinators of these circles, who had considerable experience in pharmacotherapy and group auditing. In order to compare results according to the age and sex of the patients, the two groups treated with NSAIDs were also matched (460 patients in each group). The drugs are classified under the ATC code with the volume given in defined daily doses (DDDs).

RESULTS

Marker 1 - the percentage of NSAID DDDs for recommended drugs, i.e. ibuprofen, diclofenac, indomethacin and naproxen was significantly higher for the coordinators in comparison with the high prescribers (P < 0.05). Therefore, marker 3 (drugs with questionable efficacy) and marker 5 ('me-too' drugs) show an inverse relation. Drugs with a long half-life (marker 2), high-risk drugs (marker 4) and newly marketed drugs (marker 6) were all seldomly prescribed by both groups of doctors. There was no difference between the two groups of prescribers concerning the proportion of elderly people treated with NSAIDs (marker 7). With reference to marker 8 - co-medication with anti-ulcer drugs - the coordinators treated 8.9% of NSAID patients with antacids and anti-ulcer drugs, the high prescribers, on the other hand, treated 12.2% (NS).

CONCLUSION

The markers can be easily assessed using the information obtained from drug claims and shown to each doctor personally. They call for the doctors to pay special attention to their particular drug selection. The markers can be implemented into feedback strategies of prescribing habits.