A set of four simple performance measures reflecting adherence to guidelines predicts hospitalization: a claims-based cohort study of patients with diabetes

Does the implementation of an incentive scheme increase adherence to diabetes guidelines? A retrospective cohort study of managed care enrollees

BACKGROUND

A novel incentive scheme based on a joint agreement of a large Swiss health insurance with 56 physician networks was implemented in 2018. This study evaluated the effect of its implementation on adherence to evidence-based guidelines among patients with diabetes in managed care models.

METHODS

We performed a retrospective cohort study, using health care claims data from patients with diabetes enrolled in a managed care plan (2016-2019). Guideline adherence was assessed by four evidence-based performance measures and four hierarchically constructed adherence levels. Generalized multilevel models were used to examine the effect of the incentive scheme on guideline adherence.

RESULTS

A total of 6'273 patients with diabetes were included in this study. The raw descriptive statistics showed minor improvements in guideline adherence after the implementation. After adjusting for underlying patient characteristics and potential differences between physician networks, the likelihood of receiving a test was moderately but consistently higher after the implementation of the incentive scheme for most performance measures, ranging from 18% (albuminuria: OR, 1.18; 95%-CI, 1.05-1.33) to 58% (HDL cholesterol: OR, 1.58; 95%-CI, 1.40-1.78). Full adherence was more likely after implementation of the incentive scheme (OR, 1.37; 95%-CI, 1.20-1.55), whereas level 1 significantly decreased (OR, 0.74; 95%-CI, 0.65 - 0.85). The proportions of the other adherence levels were stable.

CONCLUSION

Incentive schemes including transparency of the achieved performance may be able to improve guideline adherence in patients with diabetes and are promising to increase quality of care in this patient population.

Impact of a diabetes disease management program on guideline-adherent care, hospitalization risk and health care costs: a propensity score matching study using real-world data

OBJECTIVE

To evaluate the impact of a DMP for patients with diabetes mellitus in a Swiss primary care setting.

METHODS

In a prospective observational study, we compared diabetes patients in a DMP (intervention group; N = 538) with diabetes patients receiving usual care (control group; N = 5050) using propensity score matching with entropy balancing. Using a difference-in-difference (DiD) approach, we compared changes in outcomes from baseline (2017) to 1-year (2017/18) and to 2-year follow-up (2017/19). Outcomes included four measures for guideline-adherent diabetes care, hospitalization risk, and health care costs.

RESULTS

We identified a positive impact of the DMP on the share of patients fulfilling all measures for guideline-adherent care [DiD 2017/18: 7.2 percentage-points, p < 0.01; 2017/19: 8.4 percentage-points, p < 0.001]. The hospitalization risk was lower in the intervention group in both years, but only statistically significant in the 1-year follow-up [DiD 2017/18: - 5.7 percentage-points, p < 0.05; 2017/19: - 3.9 percentage points, n.s.]. The increase in health care costs was smaller in the intervention than in the control group [DiD 2017/18: CHF - 852; 2017/19: CHF - 909], but this effect was not statistically significant.

CONCLUSION

The DMP under evaluation seems to exert a positive impact on the quality of diabetes care, reflected in the increase in the measures for guideline-adherent care and in a reduction of the hospitalization risk in the intervention group. It also might reduce health care costs, but only a longer follow-up will show whether the observed effect persists over time.

The Effect of Financial Incentives on Quality Measures in the Treatment of Diabetes Mellitus: a Randomized Controlled Trial

BACKGROUND

Financial incentives are often used to improve quality of care in chronic care patients. However, the evidence concerning the effect of financial incentives is still inconclusive.

OBJECTIVE

To test the effect of financial incentives on quality measures (QMs) in the treatment of patients with diabetes mellitus in primary care. We incentivized a clinical QM and a process QM to test the effect of financial incentives on different types of QMs and to investigate the spill-over effect on non-incentivized QMs.

DESIGN/PARTICIPANTS

Parallel cluster randomized controlled trial based on electronic medical records database involving Swiss general practitioners (GPs). Practices were randomly allocated.

INTERVENTION

All participants received a bimonthly feedback report. The intervention group additionally received potential financial incentives on GP level depending on their performance.

MAIN MEASURES

Between-group differences in proportions of patients fulfilling incentivized QM (process QM of annual HbA1c measurement and clinical QM of blood pressure level below 140/95 mmHg) after 12 months.

KEY RESULTS

Seventy-one GPs (median age 52 years, 72% male) from 43 different practices and subsequently 3838 patients with diabetes mellitus (median age 70 years, 57% male) were included. Proportions of patients with annual HbA1c measurements remained unchanged (intervention group decreased from 79.0 to 78.3%, control group from 81.5 to 81.0%, OR 1.09, 95% CI 0.90-1.32, p = 0.39). Proportions of patients with blood pressure below 140/95 improved from 49.9 to 52.5% in the intervention group and decreased from 51.2 to 49.0% in the control group (OR 1.16, 95% CI 0.99-1.36, p = 0.06). Proportions of non-incentivized process QMs increased significantly in the intervention group.

CONCLUSION

GP level financial incentives did not result in more frequent HbA1c measurements or in improved blood pressure control. Interestingly, we could confirm a spill-over effect on non-incentivized process QMs. Yet, the mechanism of spill-over effects of financial incentives is largely unclear.

TRIAL REGISTRATION

ISRCTN13305645.

Ambulatory Healthcare Use Profiles of Patients With Diabetes and Their Association With Quality of Care: A Cross-Sectional Study

BACKGROUND

Despite the growing burden of diabetes worldwide, evidence regarding the optimal models of care to improve the quality of diabetes care remains equivocal. This study aimed to identify profiles of patients with distinct ambulatory care use patterns and to examine the association of these profiles with the quality of diabetes care.

METHODS

We performed a cross-sectional study of the baseline data of 550 non-institutionalized adults included in a prospective, community-based, cohort study on diabetes care conducted in Switzerland. Clusters of participants with distinct patterns of ambulatory healthcare use were identified using discrete mixture models. To measure the quality of diabetes care, we used both processes of care indicators (eye and foot examination, microalbuminuria screening, blood cholesterol and glycated hemoglobin measurement [HbA1c], influenza immunization, blood pressure measurement, physical activity and diet advice) and outcome indicators (12-Item Short-Form Health Survey [SF-12], Audit of Diabetes-Dependent Quality of Life [ADDQoL], Patient Assessment of Chronic Illness Care [PACIC], Diabetes Self-Efficacy Scale, HbA1c value, and blood pressure <140/90 mmHg). For each profile of ambulatory healthcare use, we calculated adjusted probabilities of receiving processes of care and estimated adjusted outcomes of care using logistic and linear regression models, respectively.

RESULTS

Four profiles of ambulatory healthcare use were identified: participants with more visits to the general practitioner [GP] than to the diabetologist and receiving concomitant podiatry care ("GP & podiatrist", n=86); participants visiting almost exclusively their GP ("GP only", n=195); participants with a substantially higher use of all ambulatory services ("High users", n=96); and participants reporting more visits to the diabetologist and less visits to the GP than other profiles ("Diabetologist first", n=173). Whereas participants belonging to the "GP only" profile were less likely to report most processes related to the quality of diabetes care, outcomes of care were relatively comparable across all ambulatory healthcare use profiles.

CONCLUSIONS

Slight differences in quality of diabetes care appear across the four ambulatory healthcare use profiles identified in this study. Overall, however, results suggest that room for improvement exists in all profiles, and further investigation is necessary to determine whether individual characteristics (like diabetes-related factors) and/or healthcare factors contribute to the differences observed between profiles.

High adherence to recommended diabetes follow-up procedures by general practitioners is associated with lower estimated cardiovascular risk

AIMS

To explore whether the general practitioners' (GPs') performance of recommended processes of care was associated with estimated risk of cardiovascular disease (CVD) and poor glycaemic control in patients with type 2 diabetes.

METHODS

A cross-sectional study from Norwegian general practice including 6015 people with type 2 diabetes <75 years old, without CVD and their 275 GPs. The GPs were split into quintiles based on each GP's average performance of six recommended processes of care. The quintiles were the exposure variable in multilevel regression models with 10-year risk of cardiovascular events estimated by NORRISK 2 (total and modifiable fraction) and poor glycaemic control (HbA_1c >69 mmol/mol (>8.5%)) as outcome variables.

RESULTS

The mean total and modifiable estimated 10-year CVD risk was 12.3% and 3.3%, respectively. Compared with patients of GPs in the highest-performing quintile, patients treated by GPs in the lowest quintile had an adjusted total and modifiable CVD risk that was 1.88 (95% CI 1.17-2.60) and 1.78 (1.14-2.41) percent point higher. This represents a relative mean difference of 16.6% higher total and 74.8% higher modifiable risk among patients of GPs in the lowest compared with the highest quintile. For patients with GPs in the lowest-performing quintile, the adjusted odds of poor glycaemic control was 1.77 (1.27-2.46) times higher than that for patients with a GP in the highest quintile.

CONCLUSIONS

We found a pattern of lower CVD risk and better glycaemic control in patients of GPs performing more recommended diabetes processes of care.

Use Of The GLYCEMIZER® Tool By General Practitioners To Meet Individual Glycated Hemoglobin Goals In Patients With Type 2 Diabetes Mellitus

Background — For general practitioners (GPs), it is often not easy to determine the individual glycated hemoglobin (HbA1c)-goal of patients with type 2 diabetes mellitus (T2DM) in order to offer them a tailored treatment and minimize side effects. Usually, they simply rely on their gut feeling. Objective — We assessed the usefulness of an easy-to-use algorithm (GLYCEMIZER®) to calculate individual HbA1c-goals and compared them with targeted (‘gut feeling’ of the GP’s) and achieved levels. Material and Methods — In this cross-sectional survey, general practitioners were asked to report anonymized data of at least 30 consecutive patients with T2DM presenting in their offices from May 1st to August 15th 2016 after obtaining informed consent. Demographic, clinical and biochemical data were used for the GLYCEMIZER® tool to calculate the individual HbA1c-goals. A statistical analysis was conducted in order to compare the calculated HbA1c-goals with targeted and achieved HbA1c-levels. Results — A total of 184 patients (mean age: 69y) were enrolled by 6 participating general practitioners from the Werdenberg-Sarganserland region in eastern Switzerland. Four patients did not meet the inclusion criteria. The overall median calculated HbA1c-goal did not differ from the targeted and achieved levels (7.1% vs. 7.0% vs. 7.1%, p=0.894). There was a significant difference between achieved and calculated HbA1c-levels in patients aged <50 (n=13, median 7.2% vs. 6.5%, p=0.014), goals not achieved) and patients aged >71 (n=85, median 6.9% vs. 7.5%, p=0.005), lower levels achieved in relation to calculated HbA1c-goals). Both in patients treated with insulin (n=44) and in patients without insulin (n=136) the achieved HbA1c-levels met the calculated goals (no insulin: 6.9% vs. 7.0%, ns; with insulin: 7.8% vs. 7.7%, ns). In regard to CKD-stages 3 and 4 the achieved HbA1c-levels were significantly lower than calculated (n= 41, median 6.9% vs. 7.6%, p=0.001). Conclusion — Calculating HbA1c-goals using the GLYCEMIZER tool is more accurate than relying on gut feeling alone, and is specifically useful in the treatment of patients with T2DM of less than 50, as well as more than 70 years of age. Furthermore, it is helpful to meet individual HbA1c-goals in patients with CKD-stages 3+.

Primary care of mothers and infants by the same or different physicians: a population-based cohort study

BACKGROUND

Different primary care delivery models exist for mothers and their infants. We examined whether primary care system performance measures differed when mother-infant dyads received primary care from the same or different providers.

METHODS

We conducted a population-based cohort study using Ontario health administrative data from 2004 to 2016. We included primiparous women and their singleton term infants and classified the primary care practitioners who provided the majority of care to the infant and the mother as concordant (same family physician for both; reference group), discordant (a different family physician for each) or pediatrician (pediatrician for the child, family physician for the mother). The primary outcome was nonobstetric maternal hospital admissions between 42 days and 2 years after delivery.

RESULTS

Among 481 721 mother-child pairs, 239 033 (49.6%) received concordant care, 114 006 (23.7%) received discordant care, and 128 682 (26.7%) received pediatrician care. Mothers in the pediatrician group were older and had greater comorbidity. Relative to concordant care, maternal nonobstetric hospital admissions occurred similarly under discordant care (adjusted odds ratio [OR] 1.00, 95% confidence interval [CI] 0.96-1.04) and in the pediatrician group (adjusted OR 0.99, 95% CI 0.95-1.02). Maternal deaths were similar under discordant care (adjusted OR 1.00, 95% CI 0.62-1.63) but lower in the pediatrician group (adjusted OR 0.55, 95% CI 0.34-0.89). Maternal primary care visits were lower in both the discordant group (adjusted relative risk [RR] 0.68, 95% CI 0.68-0.69) and the pediatrician group (adjusted RR 0.75, 95% CI 0.75-0.76). Healthy children were more likely to miss the enhanced 18-month well-baby visit under discordant care (adjusted OR 1.06, 95% CI 1.03-1.09) but less likely to miss this visit under pediatrician care (adjusted OR 0.47, 95% CI 0.46-0.49).

INTERPRETATION

Concordant care provided to a new mother and her infant by the same family physician was not associated with better primary care health system performance. The reason that pediatric primary care is associated with better maternal and child outcomes remains to be determined.

Association Between Routine Nephropathy Monitoring and Subsequent Change in Estimated Glomerular Filtration Rate in Patients With Diabetes Mellitus: A Japanese Non-Elderly Cohort Study

Backgrounds

Current guidelines recommend routine nephropathy monitoring, including microalbuminuria or proteinuria testing, for people with diabetes mellitus; however, its effect in terms of preserving renal function remains unclear. We conducted this study to examine the impact of routine nephropathy monitoring on subsequent changes in estimated glomerular filtration rate.

Methods

We retrospectively identified non-elderly individuals with diabetes mellitus based on the prescription of hypoglycemic agents from a large Japanese database (JMDC, Tokyo, Japan) of screening for lifestyle diseases linked with administrative claims data. We collected data on baseline characteristics including age, sex, comorbidity, and laboratory data. We then examined the association between routine nephropathy monitoring results and change in estimated glomerular filtration rate using a propensity-score inverse probability of treatment weighting method.

Results

Among 1,602 individuals who started taking hypoglycemic agents between 2005 and 2016, 102 (6.0%) underwent routine nephropathy monitoring during the first year of medication for diabetes mellitus. After adjusting for multiple confounding factors, there was no significant difference in subsequent estimated glomerular filtration rate changes between individuals with and without routine nephropathy monitoring (difference in percent change 0.11; 95% confidence interval −2.74 to 2.95).

Conclusion

Routine nephropathy monitoring was not associated with preserved renal function. Current recommendations for the universal application of nephropathy monitoring may have limited value to prevent renal dysfunction in non-elderly individuals with diabetes mellitus.

Quality performance and associated factors in Swiss diabetes care - A cross-sectional study

INTRODUCTION

Quality indicators and pay-for-performance schemes aim to improve processes and outcomes in clinical practice. However, general practitioner and patient characteristics influence quality indicator performance. In Switzerland, no data on the pay-for-performance approach exists and the use of quality indicators has been marginal. The aim of this study was to describe quality indicator performance in diabetes care in Swiss primary care and to analyze associations of practice, general practitioner and patient covariates with quality indicator performance.

METHODS

For this cross-sectional study, we used medical routine data from an electronic medical record database. Data from 71 general practitioners and all their patients with diabetes were included. Starting in July 2018, we retrieved 12-month retrospective data about practice, general practitioner and patient characteristics, laboratory values, comorbidities and co-medication. Based on this data, we assessed quality indicator performance of process and intermediate outcomes for glycated hemoglobin, blood pressure, cholesterol and associations of practice, general practitioner and patient characteristics with individual and cumulative quality indicator performance. We calculated odds ratios (OR) and 95% confidence intervals (CI) using regression methods.

RESULTS

We assessed 3,383 patients with diabetes (57% male, mean age 68.3 years). On average, patients fulfilled 3.56 (standard deviation: 1.89) quality indicators, whereas 17.2% of the patients fulfilled all six quality indicators. On practice and general practitioner level, we found no associations with cumulative quality indicator performance. On patient level, gender (ref = male) (OR: 0.83, CI: 0.78-0.88), number of treating general practitioners (OR: 0.94, CI: 0.91-0.97), number of comorbidities (OR: 1.43, CI: 1.38-1.47) and number of consultations (OR: 1.02, CI: 1.02-1.02) were associated with cumulative quality indicator performance.

CONCLUSION

The influence of practice, general practitioner and patient characteristics on quality indicator performance was surprisingly small and room for improvement in quality indicator performance of Swiss general practitioners seems to exist in diabetes care.

Evidence-based quality indicators for primary healthcare in association with the risk of hospitalisation: a population-based cohort study in Switzerland

OBJECTIVES

The quality of ambulatory care in Switzerland is widely unknown. Therefore, this study aimed to evaluate the recently proposed quality indicators (QIs) based on a nationwide healthcare claims database and determine their association with the risk of subsequent hospitalisation at patient-level.

DESIGN

Retrospective cohort study.

SETTING

Inpatient and outpatient claims data of a large health insurance in Switzerland covering all regions and population strata.

PARTICIPANTS

520 693 patients continuously insured during 2015 and 2016.

MEASURES

A total of 24 QIs were obtained by adapting the existing instruments to the Swiss national context and measuring at patient-level. The association between each QI and hospitalisation in the subsequent year was assessed using multiple logistic regression models.

RESULTS

The proportion of patients with good adherence to QIs was high for the secondary prevention of diabetes and myocardial infarction (glycated haemoglobin (HbA1c) control, 89%; aspirin use, 94%) but relatively low for polypharmacy (53%) or using potentially inappropriate medications (PIMs) in the elderly (PIM, 33%). Diabetes-related indicators such as the HbA1c control were significantly associated with a lower risk of hospitalisation (OR, 0.87; 95% CI, 0.80 to 0.95), whereas the occurrence of polypharmacy and PIM increased the risk of hospitalisation in the following year (OR, 1.57/1.08; 95% CI, 1.51 to 1.64/1.05 to 1.12).

CONCLUSIONS

This is the first study to evaluate the recently presented QIs in Switzerland using nationwide real-life data. Our study suggests that the quality of healthcare, as measured by these QIs, varied. The majority of QIs, in particular QIs reflecting chronic care and medication use, are considered beneficial markers of healthcare quality as they were associated with reduced risk of hospitalisation in the subsequent year. Results from this large practical test on real-life data show the feasibility of these QIs and are beneficial in selecting the appropriate QIs for healthcare implementation in general practice.

Adherence to medical follow-up recommendations reduces hospital admissions: Evidence from diabetic patients in France

The aim of this study was to document the extent to which diabetic patients who adhered to required medical follow-ups in France experienced reduced hospital admissions over time. The main assumption was that enhanced monitoring and follow-up of diabetic patients in the primary care setting could be a substitute for hospital use. Using longitudinal claim data of diabetic patients between 2010 and 2015 from MGEN, a leading mutuelle insurance company in France, we estimated a dynamic logit model with lagged measures of the quality of adherence to eight medical follow-up recommendations. This model allowed us to disentangle follow-up care in hospitals from other forms of inpatient care that could occur simultaneously. We found that a higher adherence to medical guidance is associated with a lower probability of hospitalization and that the take-up of each of the eight recommendations may help reduce the rates of hospital admission. The reasons for the variation in patient adherence and implications for health policy are discussed.

Exploring geographic variation of and influencing factors for utilization of four diabetes management measures in Swiss population using claims data

INTRODUCTION

Four strongly recommended diabetes management measures are biannual glycated hemoglobin (HbA1c) testing, annual eye examination, kidney function examination, and low-density lipoprotein (LDL) testing in patients below 75 years. We aimed to describe regional variation in the utilization of the four measures across small regions in Switzerland and to explore potential influencing factors.

RESEARCH DESIGN AND METHODS

We conducted a cross-sectional study of adult patients with drug-treated diabetes in 2014 using claims data. Four binary outcomes represented adherence to the recommendations. Possible influencing factors included sociodemographics, health insurance preferences, and clinical characteristics. We performed multilevel modeling with Medstat regions as the higher level. We calculated the median odds ratio (MOR) and checked spatial autocorrelation in region level residuals using Moran's I statistic. When significant, we further conducted spatial multilevel modeling.

RESULTS

Of 49 198 patients with diabetes (33 957 below 75 years), 69.6% had biannual HbA1c testing, 44.3% each had annual eye examination and kidney function examination, and 55.5% of the patients below 75 years had annual LDL testing. The effects of health insurance preferences were substantial and consistent. Having any supplementary insurance (ORs across measures were between 1.08 and 1.28), having supplementary hospital care insurance (1.08-1.30), having chosen a lower deductible level (eg, SFr2500 compared with SFr300: 0.57-0.69), and having chosen a managed care model (1.04-1.17) were positively associated with recommendations adherence. The MORs (1.27-1.33) showed only moderate unexplained variation, and we observed inconsistent spatial patterns of unexplained variation across the four measures.

CONCLUSION

Our findings indicate that the uptake of strongly recommended measures in diabetes management could possibly be optimized by providing further incentives to patients and care providers through insurance scheme design. The absence of marked regional variation implies limited potential for improvement by targeted regional intervention, while provider-specific promotion may be more impactful.

Measuring diabetes guideline adherence with claims data: systematic construction of indicators and related challenges

OBJECTIVES

Indicators of guideline adherence are frequently used to examine the appropriateness of healthcare services. Only some potential indicators are actually usable for research with routine administrative claims data, potentially leading to a biased selection of research questions. This study aimed at developing a systematic approach to extract potential indicators from clinical practice guidelines (CPG), evaluate their feasibility for research with claims data and assess how the extracted set reflected different types of healthcare services. Diabetes mellitus (DM), Swiss national guidelines and health insurance claims data were analysed as a model case.

METHODS

CPG for diabetes patients were retrieved from the Swiss Endocrinology and Diabetes Society website. Recommendation statements involving a specific healthcare intervention for a defined patient population were translated into indicators of guideline adherence. Indicators were classified according to disease stage and healthcare service type. We assessed for all indicators whether they could be analysed with Swiss mandatory health insurance administrative claims data.

RESULTS

A total of 93 indicators were derived from 15 CPG, representing all sectors of diabetes care. For 63 indicators, the target population could not be identified using claims data only. For 67 indicators, the intervention could not be identified. Nine (10%) of all indicators were feasible for research with claims data (three addressed gestational diabetes and screening, five screening for complications and one glucose measurement). Some types of healthcare services, eg, management of risk factors, treatment of the disease and secondary prevention, lacked corresponding indicators feasible for research.

CONCLUSIONS

Our systematic approach could identify a number of indicators of healthcare service utilisation, feasible for DM research with Swiss claims data. Some areas of healthcare were covered less well. The approach could be applied to other diseases and countries, helping to identify the potential bias in the selection of indicators and optimise research.

Evidence-based indicators for the measurement of quality of primary care using health insurance claims data in Switzerland: results of a pragmatic consensus process

BACKGROUND

The level of quality of care of ambulatory services in Switzerland is almost completely unknown. By adapting existing instruments to the Swiss national context, the present project aimed to define quality indicators (QI) for the measurement of quality of primary care for use on health insurance claims data. These data are pre-existing and available nationwide which provides an excellent opportunity for their use in the context of health care quality assurance.

METHODS

Pragmatic 6-step process based on informal consensus. Potential QI consisted of recommendations extracted from internationally accepted medical practice guidelines and pre-existing QI for primary care. An independent interdisciplinary group of experts rated potential QI based on explicit criteria related to evidence, relevance for Swiss public health, and controllability in the Swiss primary care context. Feasibility of a preliminary set of QI was tested using claims data of persons with basic mandatory health insurance with insurance at one of the largest Swiss health insurers. This test built the basis for expert consensus on the final set of QI.

RESULTS

Of 49 potential indicators, 23 were selected for feasibility testing based on claims data. The expert group consented a final set of 24 QI covering the domains general aspects/ efficiency (7 QI), drug safety (2), geriatric care (4), respiratory disease (2), diabetes (5) and cardiovascular disease (4).

CONCLUSIONS

The present project provides the first nationwide applicable explicit evidence-based criteria to measure quality of care of ambulatory primary care in Switzerland. The set intends to increase transparency related to quality and variance of care in Switzerland.

[Routine Data in Health Services Research: an Underused Resource]

Zusammenfassung. Routinedaten entstehen als Nebenprodukt des normalen Betriebsalltags. Diese ohne Zusatzaufwand massenhaft generierten Daten lassen sich mit ökonomischen oder im Falle der Medizin auch mit gesundheitlichen Fragestellungen auswerten. Durch die nicht-kontrollierte Art der Sammlung von Routinedaten sowie aufgrund der uneinheitlichen Handhabung von elektronischen Krankengeschichten ergeben sich aber je nach Fragestellung Grenzen der Aussagekraft von Resultaten. Dieser Artikel widmet sich den Fragen rund um Herkunft, Verarbeitung, Interpretation und Nutzen von Routinedaten. Dabei werden auch Machine Learning und Big Data kritisch in den Kontext gebracht, sowie die Aspekte Datenschutz und Ethik. Hinsichtlich des Schweizer Gesundheitssystems zeigt der Artikel die notwendigen Voraussetzungen, damit das Potential von Routinedaten auch hierzulande zugunsten einer besseren medizinischen Versorgung genutzt werden kann.