Characteristics Associated With Telemonitoring Use Among Patients With Chronic Heart Failure: Retrospective Cohort Study.

BACKGROUND: Chronic heart failure (HF) is a chronic disease affecting more than 64 million people worldwide, with an increasing prevalence and a high burden on individual patients and society. Telemonitoring may be able to mitigate some of this burden by increasing self-management and preventing use of the health care system. However, it is unknown to what degree telemonitoring has been adopted by hospitals and if the use of telemonitoring is associated with certain patient characteristics. Insight into the dissemination of this technology among hospitals and patients may inform strategies for further adoption. OBJECTIVE: We aimed to explore the use of telemonitoring among hospitals in the Netherlands and to identify patient characteristics associated with the use of telemonitoring for HF. METHODS: We performed a retrospective cohort study based on routinely collected health care claim data in the Netherlands. Descriptive analyses were used to gain insight in the adoption of telemonitoring for HF among hospitals in 2019. We used logistic multiple regression analyses to explore the associations between patient characteristics and telemonitoring use. RESULTS: Less than half (31/84, 37%) of all included hospitals had claims for telemonitoring, and 20% (17/84) of hospitals had more than 10 patients with telemonitoring claims. Within these 17 hospitals, a total of 7040 patients were treated for HF in 2019, of whom 5.8% (409/7040) incurred a telemonitoring claim. Odds ratios (ORs) for using telemonitoring were higher for male patients (adjusted OR 1.90, 95% CI 1.50-2.41) and patients with previous hospital treatment for HF (adjusted OR 1.76, 95% CI 1.39-2.24). ORs were lower for higher age categories and were lowest for the highest age category, that is, patients older than 80 years (OR 0.30, 95% CI 0.21-0.44) compared to the reference age category (18-59 years). Socioeconomic status, degree of multimorbidity, and excessive polypharmacy were not associated with the use of telemonitoring. CONCLUSIONS: The use of reimbursed telemonitoring for HF was limited up to 2019, and our results suggest that large variation exists among hospitals. A lack of adoption is therefore not only due to a lack of diffusion among hospitals but also due to a lack of scaling up within hospitals that already deploy telemonitoring. Future studies should therefore focus on both kinds of adoption and how to facilitate these processes. Older patients, female patients, and patients with no previous hospital treatment for HF were less likely to use telemonitoring for HF. This shows that some patient groups are not served as much by telemonitoring as other patient groups. The underlying mechanism of the reported associations should be identified in order to gain a deeper understanding of telemonitoring use among different patient groups.

Privacy-preserving dataset combination and Lasso regression for healthcare predictions.

BACKGROUND: Recent developments in machine learning have shown its potential impact for clinical use such as risk prediction, prognosis, and treatment selection. However, relevant data are often scattered across different stakeholders and their use is regulated, e.g. by GDPR or HIPAA. As a concrete use-case, hospital Erasmus MC and health insurance company Achmea have data on individuals in the city of Rotterdam, which would in theory enable them to train a regression model in order to identify high-impact lifestyle factors for heart failure. However, privacy and confidentiality concerns make it unfeasible to exchange these data. METHODS: This article describes a solution where vertically-partitioned synthetic data of Achmea and of Erasmus MC are combined using Secure Multi-Party Computation. First, a secure inner join protocol takes place to securely determine the identifiers of the patients that are represented in both datasets. Then, a secure Lasso Regression model is trained on the securely combined data. The involved parties thus obtain the prediction model but no further information on the input data of the other parties. RESULTS: We implement our secure solution and describe its performance and scalability: we can train a prediction model on two datasets with 5000 records each and a total of 30 features in less than one hour, with a minimal difference from the results of standard (non-secure) methods. CONCLUSIONS: This article shows that it is possible to combine datasets and train a Lasso regression model on this combination in a secure way. Such a solution thus further expands the potential of privacy-preserving data analysis in the medical domain.

Prediction of long-term hospitalisation and all-cause mortality in patients with chronic heart failure on Dutch claims data: a machine learning approach.

BACKGROUND: Accurately predicting which patients with chronic heart failure (CHF) are particularly vulnerable for adverse outcomes is of crucial importance to support clinical decision making. The goal of the current study was to examine the predictive value on long term heart failure (HF) hospitalisation and all-cause mortality in CHF patients, by exploring and exploiting machine learning (ML) and traditional statistical techniques on a Dutch health insurance claims database. METHODS: Our study population consisted of 25,776 patients with a CHF diagnosis code between 2012 and 2014 and one year and three years follow-up HF hospitalisation (1446 and 3220 patients respectively) and all-cause mortality (2434 and 7882 patients respectively) were measured from 2015 to 2018. The area under the receiver operating characteristic (ROC) curve (AUC) was calculated after modelling the data using Logistic Regression, Random Forest, Elastic Net regression and Neural Networks. RESULTS: AUC rates ranged from 0.710 to 0.732 for 1-year HF hospitalisation, 0.705-0.733 for 3-years HF hospitalisation, 0.765-0.787 for 1-year mortality and 0.764-0.791 for 3-years mortality. Elastic Net performed best for all endpoints. Differences between techniques were small and only statistically significant between Elastic Net and Logistic Regression compared with Random Forest for 3-years HF hospitalisation. CONCLUSION: In this study based on a health insurance claims database we found clear predictive value for predicting long-term HF hospitalisation and mortality of CHF patients by using ML techniques compared to traditional statistics.

Outcome measures for oral health based on clinical assessments and claims data: feasibility evaluation in practice.

BACKGROUND: It is well known that treatment variation exists in oral healthcare, but the consequences for oral health are unknown as the development of outcome measures is still in its infancy. The aim of this study was to identify and develop outcome measures for oral health and explore their performance using health insurance claims records and clinical data from general dental practices. METHODS: The Dutch healthcare insurance company Achmea collaborated with researchers, oral health experts, and general dental practitioners (GDPs) in a proof of practice study to test the feasibility of measures in general dental practices. A literature search identified previously described outcome measures for oral healthcare. Using a structured approach, identified measures were (i) prioritized, adjusted and added to after discussion and then (ii) tested for feasibility of data collection, their face validity and discriminative validity. Data sources were claims records from Achmea, clinical records from dental practices, and prospective, pre-determined clinical assessment data obtained during routine consultations. RESULTS: In total eight measures (four on dental caries, one on tooth wear, two on periodontal health, one on retreatment) were identified, prioritized and tested. The retreatment measure and three measures for dental caries were found promising as data collection was feasible, they had face validity and discriminative validity. Deployment of these measures demonstrated variation in clinical practices of GDPs. Feedback of this data to GDPs led to vivid discussions on best practices and quality of care. The measure 'tooth wear' was not considered sufficiently responsive; 'changes in periodontal health score' was considered a controversial measure. The available data for the measures 'percentage of 18-year-olds with no tooth decay' and 'improvement in gingival bleeding index at reassessment' was too limited to provide accurate estimates per dental practice. CONCLUSIONS: The evaluated measures 'time to first restoration', 'distribution of risk categories for dental caries', 'filled-and-missing score' and 'retreatment after restoration', were considered valid and relevant measures and a proxy for oral health status. As such, they improve the transparency of oral health services delivery that can be related to oral health outcomes, and with time may serve to improve these oral health outcomes.

Measuring Healthcare Providers' Performances Within Managed Competition Using Multidimensional Quality and Cost Indicators.

BACKGROUND AND OBJECTIVES: The Dutch healthcare system is in transition towards managed competition. In theory, a system of managed competition involves incentives for quality and efficiency of provided care. This is mainly because health insurers contract on behalf of their clients with healthcare providers on, potentially, quality and costs. The paper develops a strategy to comprehensively analyse available multidimensional data on quality and costs to assess and report on the relative performance of healthcare providers within managed competition. DATA AND METHODS: We had access to individual information on 2409 clients of 19 Dutch diabetes care groups on a broad range of (outcome and process related) quality and cost indicators. We carried out a cost-consequences analysis and corrected for differences in case mix to reduce incentives for risk selection by healthcare providers. RESULTS AND CONCLUSION: There is substantial heterogeneity between diabetes care groups' performances as measured using multidimensional indicators on quality and costs. Better quality diabetes care can be achieved with lower or higher costs. Routine monitoring using multidimensional data on quality and costs merged at the individual level would allow a systematic and comprehensive analysis of healthcare providers' performances within managed competition.

Predicting inadequate bowel preparation for colonoscopy in participants receiving split-dose bowel preparation: development and validation of a prediction score.

BACKGROUND: Adequate bowel preparation is important for optimal colonoscopy. It is important to identify patients at risk for inadequate bowel preparation because this allows taking precautions in this specific group. OBJECTIVE: To develop a prediction score to identify patients at risk for inadequate bowel preparation who may benefit from an intensified bowel cleansing regimen. DESIGN: Patient and colonoscopy data were prospectively collected, whereas clinical data were retrospectively collected for a total of 1996 colonoscopies in participants who received split-dose bowel preparation. Multivariate logistic regression analyses were conducted in a random two-thirds of the cohort to develop a prediction model. Validation and evaluation of the discriminative power of the prediction model were performed within the remaining one-third of the cohort. SETTING: Four centers, including one academic and three medium-to-large size nonacademic centers. PATIENTS: Consecutive colonoscopies in November and December 2012. Mean age was 57.3 ± 15.9 years, 45.8% were male and indications for colonoscopy were screening and/or surveillance (27%), abdominal symptoms and/or blood loss and/or anemia (60%), inflammatory bowel disease (9%), and others (4%). INTERVENTIONS: Colonoscopy. MAIN OUTCOME MEASUREMENTS: Inadequate bowel preparation defined as Boston Bowel Preparation Scale score <6. RESULTS: A total of 1331 colonoscopies were included in the development cohort, of which 172 (12.9%) had an inadequate bowel preparation. Independent factors included in the prediction model were American Society of Anesthesiologists Physical Status Classification System score ≥3, use of tricyclic antidepressants, use of opioids, diabetes, chronic constipation, history of abdominal and/or pelvic surgery, history of inadequate bowel preparation, and current hospitalization. The discriminative ability of the scale was good, with an area under the curve of 0.77 in the validation cohort. LIMITATIONS: Study design partially retrospective, no data on patient compliance. CONCLUSION: We developed a validated, easy-to-use prediction scale that can be used to identify subjects with an increased risk of inadequate bowel preparation with good accuracy.

Comparison of cecal intubation and adenoma detection between hospitals can provide incentives to improve quality of colonoscopy.

BACKGROUND AND STUDY AIMS: Cecal intubation rate (CIR) and adenoma detection rate (ADR) have been found to be inversely associated with the occurrence of post-colonoscopy colorectal cancer. Depicting differences in CIR and ADR between hospitals could provide incentives for quality improvement. The aim of this study was to compare quality parameters of routine colonoscopies between seven hospitals in The Netherlands in order to determine the extent to which possible differences were attributable to procedural and institutional factors. PATIENTS AND METHODS: Consecutive patients undergoing colonoscopy were prospectively included between November 2012 and January 2013 at two academic and five nonacademic hospitals. Patients with inflammatory bowel disease or hereditary colorectal cancer syndromes were excluded. Main outcome measures were CIR and ADR. RESULTS: A total of 3129 patients were included (mean age 59 ±â€Š15 years; 45.5 % male). The majority of patients (86.2 %) had a Boston Bowel Preparation Scale (BBPS) score ≥ 6. Overall CIR was 94.8 %, ranging from 89.4 % to 99.2 % between hospitals. After adjustment for case mix (age, sex, American Society of Anesthesiologists score, and indication for colonoscopy), factors associated with CIR were hospital and a BBPS score ≥ 6. Overall ADR was 31.8 % and varied between hospitals, ranging from 24.8 % to 46.8 %. Independent predictors for ADR were hospital, BBPS score ≥ 6, and cecal intubation. By combining CIR and ADR for each hospital, a colonoscopy quality indicator (CQI) was developed, which can be used by hospitals to stimulate quality improvement. CONCLUSION: Differences in the quality of colonoscopy between hospitals can be demonstrated using CIR and ADR. As both indicators are affected by institution and bowel preparation, a comparison between hospitals based on the newly developed CQI could assist in further improving the quality of colonoscopy.

Effectiveness of pharmacotherapy in behavioural therapeutic smoking cessation programmes.

BACKGROUND: In 2011, pharmacotherapy as a part of smoking cessation treatment was reimbursed through the basic health insurance in the Netherlands. We examine the (cost)-effectiveness of pharmacotherapy added to behavioural therapy. METHODS: An observational study was conducted using data from the suppliers of the smoking cessation programmes together with information on costs from health insurance company Achmea. National suppliers, general practitioners and healthcare centres offered four different programmes. (i) Behavioural support (=therapy); (ii) Behavioural support combined with nicotine replacement therapy (NRT); (iii) Behavioural support combined with smoking cessation aids (=medication) (SCA); (iv) Behavioural support combined with NRT and SCA. The primary independent variable was the programme type, and the primary outcome was whether someone quitted smoking. To examine the effectiveness of the different programmes logistic regression and logistic multilevel analyses were performed. Bootstrapping was used to evaluate cost-effectiveness. RESULTS: The results indicate that behavioural support combined with SCA has more quitters than the reference programme of behavioural support alone, and it also seems the most cost-effective programme for general practitioners and healthcare centres. Behavioural therapy combined with NRT had also more quitters, although the difference with the reference programme was smaller. CONCLUSION: Behavioural support combined with SCA seems the most successful programme. However, as we performed an observational study, firm conclusions about the differences in effectiveness between the programme types cannot be made. Future research should consider the type of smoker (smoking history, amount of cigarettes per day).

An examination of the risk of periodontitis for nonfatal cardiovascular diseases on the basis of a large insurance claims database.

OBJECTIVES: Although many studies have reported a higher risk of atherosclerotic cardiovascular diseases (ACVD) in people with periodontitis (PD), this has been tested in a few large-scale population-based studies with a longitudinal design. The aim of this study was to investigate whether people with PD status have an increased risk of a nonfatal ACVD event compared to people without PD status. METHODS: A cohort of 1.2 million participants from a healthcare insurance claims database was studied longitudinally for a period of 8 years. PD status was derived from PD-related insurance claims and ACVD status from ACVD-related insurance claims. Person-time at risk (PTAR) was calculated from the start of follow-up (01 January 2007) for participants with and without PD status until ACVD or event-free censoring (31 December 2014). Time-dependent Cox proportional hazard models were used to calculate the hazard ratio (HR) and to adjust for shared risk factors (age, sex, socioeconomic position and diabetes mellitus). RESULTS: The prevalence of PD was 20.1%, and the cumulative incidence of nonfatal ACVD events was 7.5%. The univariable and multivariable analyses revealed a limited risk of ACVD for participants with PD status (HR: 1.12; 95% CI 1.10-1.14, HR: 1.06; 95% CI 1.04-1.08, respectively). A subgroup analysis of participants ≤35 and > 35 years of age showed that those ≤35 years of age with PD status had a higher ACVD risk (univariable HR: 1.20; 95% CI 1.05-1.37, multivariable HR: 1.21; 95% CI 1.05-1.39). ACVD risk was not increased in participants >35 years of age with PD status (univariable HR: 0.92; 95% CI 0.91-0.94, multivariable HR: 0.96; 95% CI 0.94-0.98). CONCLUSIONS: This study based on a healthcare insurance cohort shows that PD can hardly be regarded as a risk factor for nonfatal ACVD. The increased risk is of minor size, and therefore, the proposed role of PD in the development of ACVD events should be reconsidered. Possibly PD plays a role as a risk factor in younger people due to overlapping genetic risk factors of ACVD and a more aggressive course of PD.

Does Centralization of Radical Prostatectomy Reduce the Incidence of Postoperative Urinary Incontinence?

Background: On the basis of previous analyses of the incidence of urinary incontinence (UI) after radical prostatectomy (RP), the hospital RP volume threshold in the Netherlands was gradually increased from 20 per year in 2017, to 50 in 2018 and 100 from 2019 onwards. Objective: To evaluate the impact of hospital RP volumes on the incidence and risk of UI after RP (RP-UI). Design setting and participants: Patients who underwent RP during 2016-2020 were identified in the claims database of the largest health insurance company in the Netherlands. Incontinence was defined as an insurance claim for ≥1 pads/d. Outcome measurements and statistical analysis: The relationship between hospital RP volume (HV) and RP-UI was assessed via multivariable analysis adjusted for age, comorbidity, postoperative radiotherapy, and lymph node dissection. Results and limitations: RP-UI incidence nationwide and by RP volume category did not decrease significantly during the study period, and 5-yr RP-UI rates varied greatly among hospitals (19-85%). However, low-volume hospitals (≤120 RPs/yr) had a higher percentage of patients with RP-UI and higher variation in comparison to high-volume hospitals (>120 RPs/yr). In comparison to hospitals with low RP volumes throughout the study period, the risk of RP-UI was 29% lower in hospitals shifting from the low-volume to the high-volume category (>120 RPs/yr) and 52% lower in hospitals with a high RP volume throughout the study period (>120 RPs/yr for 5 yr). Conclusions: A focus on increasing hospital RP volumes alone does not seem to be sufficient to reduce the incidence of RP-UI, at least in the short term. Measurement of outcomes, preferably per surgeon, and the introduction of quality assurance programs are recommended. Patient summary: In the Netherlands, centralization of surgery to remove the prostate (RP) because of cancer has not yet improved the occurrence of urinary incontinence (UI) after surgery. Hospitals performing more than 120 RP operations per year had better UI outcomes. However, there was a big difference in UI outcomes between hospitals.

Personalizing the Use of a Intermittently Scanned Continuous Glucose Monitoring (isCGM) Device in Individuals With Type 1 Diabetes: A Cost-Effectiveness Perspective in the Netherlands (FLARE-NL 9).

AIMS: Intermittently scanned continuous glucose monitoring (isCGM) is a method to monitor glucose concentrations without using a finger prick. Among persons with type 1 diabetes (T1D), isCGM results in improved glycemic control, less disease burden and improved health-related quality of life (HRQoL). However, it is not clear for which subgroups of patients isCGM is cost-effective. We aimed to provide a real-world cost-effectiveness perspective. METHODS: We used clinical data from a 1-year nationwide Dutch prospective observational study (N = 381) and linked these to insurance records. Health-related quality of life was assessed with the EQ-5D-3L questionnaire. Individuals were categorized into 4 subgroups: (1) frequent hypoglycemic events (58%), (2) HbA1c > 70 mmol/mol (8.5%) (19%), (3) occupation that requires avoiding finger pricks and/or hypoglycemia (5%), and (4) multiple indications (18%). Comparing costs and outcomes 12 months before and after isCGM initiation, incremental cost-effectiveness ratios (ICERs) were calculated for the total cohort and each subgroup from a societal perspective (including healthcare and productivity loss costs) at the willingness to pay of €50,000 per quality-adjusted life year (QALY) gained. RESULTS: From a societal perspective, isCGM was dominant in all subgroups (ie higher HRQoL gain with lower costs) except for subgroup 1. From a healthcare payer perspective, the probabilities of isCGM being cost-effective were 16%, 9%, 30%, 98%, and 65% for the total cohort and subgroup 1, 2, 3, and 4, respectively. Most sensitivity analyses confirmed these findings. CONCLUSIONS: Comparing subgroups of isCGM users allows to prioritize them based on cost-effectiveness. The most cost-effective subgroup was occupation-related indications, followed by multiple indications, high HbA1c and the frequent hypoglycemic events subgroups. However, controlled studies with larger sample size are needed to draw definitive conclusions.

Impact of sex differences in co-morbidities and medication adherence on outcome in 25 776 heart failure patients.

AIMS: Health insurance claims (HIC) databases in the Netherlands capture unselected patient populations, which makes them suitable for epidemiological research on sex differences. Based on a HIC database, we aimed to reveal sex differences in heart failure (HF) outcomes, with particular focus on co-morbidities and medication. METHODS AND RESULTS: The Achmea HIC database included 14 517 men and 11 259 (45%) women with a diagnosis treatment code for chronic HF by January 2015. We related their sex, co-morbidities, and medication adherence (medication possession rate >0.8) with the primary endpoint (PE) of all-cause mortality or HF admission during a median follow-up of 3.3 years, using Cox regression. Median age of men and women was 72 and 76 years, respectively. Prevalence of co-morbidities and use of disease-modifying drugs was higher in men; however, medication adherence was similar. At the end of follow-up, 35.1% men and 31.8% women had reached the PE. The adjusted hazard ratio for men was 1.25 (95% confidence interval: 1.19-1.30). A broad range of co-morbidities was associated with the PE. Overall, these associations were stronger in women than in men, particularly for renal insufficiency, chronic obstructive pulmonary disease/asthma, and diabetes. Non-adherence to disease-modifying drugs was related with a higher incidence of the PE, with similar effects between sexes. CONCLUSIONS: In a representative sample of the Dutch population, as captured in a HIC database, men with chronic HF had a 25% higher incidence of death or HF admission than women. The impact of co-morbidities on the outcome was sex dependent, while medication adherence was not.

Once bitten, twice shy? Lessons learned from an experiment to liberalize price regulations for dental care.

In 2012, The Netherlands established the so-called "free market experiment", which allowed providers of dental care to set the prices for their dental services themselves. The introduction of market mechanisms is intended to improve the quality of care and to contribute to cost containment, but increasing health expenditures for citizens have been observed in this context. Using large-volume health insurance claims data and exploiting the 2012 experiment in Dutch dental care, we identified the effects of a liberalization of service prices. Using pooled regression with individual fixed effects, we analyzed changes in utilization patterns of prevention-oriented dental services in response to the experiment as well as the elasticities in demand in response to variations in out-of-pocket (OOP) prices. We found substantial increases in prices and patients' OOP contributions for dental services following the liberalization with differences in increases between types of services. In response to the experiment, the proportion of treatment sessions containing preventive-oriented services decreased significantly by 3.4% among adults and by 5.3% for children and adolescents. Estimates of short-run price elasticities of demand for different services point towards differences in price sensitivity. One potential explanation for the observed variations in prices and utilization could be different extents of asymmetric information for first-stage and follow-on services. Price liberalization seems to have affected the composition of treatment sessions towards a decreasing use of preventive services, suggesting a shift in the reason for seeing a dental care provider from a regular-preventive perspective to a symptom-based restorative approach.

Effect of periodontal treatment on diabetes-related healthcare costs: a retrospective study.

INTRODUCTION: Periodontitis has been considered a sixth complication of diabetes. The aim of this study was to assess the impact of periodontal treatment on diabetes-related healthcare costs in patients with diabetes. RESEARCH DESIGN AND METHODS: A retrospective analysis was done, exploiting unique and large-scale claims data of a Dutch health insurance company. Data were extracted for a cohort of adults who had been continuously insured with additional dental coverage for the years 2012-2018. Individuals with at least one diabetes-related treatment claim in 2012 were included for analysis. A series of panel data regression models with patient-level fixed effects were estimated to assess the impact of periodontal treatment on diabetes-related healthcare costs. RESULTS: A total of 41 598 individuals with diabetes (age range 18-100 years; 45.7% female) were included in the final analyses. The median diabetes-related healthcare costs per patient in 2012 were €38.45 per quarter (IQR €11.52-€263.14), including diagnoses, treatment, medication and hospitalization costs. The fixed effect models showed €12.03 (95% CI -€15.77 to -€8.29) lower diabetes-related healthcare costs per quarter of a year following periodontal treatment compared with no periodontal treatment. CONCLUSIONS: Periodontitis, a possible complication of diabetes, should receive appropriate attention in diabetes management. The findings of this study provide corroborative evidence for reduced economic burdens due to periodontal treatment in patients with diabetes.

The effect of reinsuring a deductible on pharmaceutical spending: A Dutch case study on low-income people.

Many countries have cost sharing schemes in health insurance to control health care expenditures. The Dutch basic health insurance includes a mandatory deductible of currently 385 euros per adult per year. To avoid affordability problems, several municipalities offer a group contract for low-income people in which the mandatory deductible is 'reinsured'. More specifically, this means that out-of-pocket spending under the deductible is covered by supplementary insurance. By comparing groups with and without the reinsurance option, this study examines whether low-income people are price-sensitive when it comes to pharmaceutical spending. We use a unique dataset from a Dutch health insurer with anonymized individual insurance claims for the period 2014-2017. The data allows for a clean difference-in-difference analysis as it contains both municipalities without reinsurance and municipalities that introduced reinsurance on January 1st 2017. We find that the introduction of reinsurance led to a statistically significant increase in pharmaceutical spending of 16% in the first quarter of 2017 and 7% in the second quarter. For the second half of 2017 the effect is small and not statistically significant. This study adds to the evidence that low-income people are indeed price-sensitive when it comes to pharmaceutical spending.

Do independent treatment centers offer more value than general hospitals? The case of cataract care.

OBJECTIVE: To identify differences between independent treatment centers (ITCs) and general hospitals (GHs) regarding costs, quality of care, and efficiency. DATA SOURCES: Anonymous claims data (2013-2015) were used. We also obtained quality indicators from a semipublic platform. STUDY DESIGN: This study uses a comparative multilevel analysis, controlling for case mix, to evaluate the performance of ITCs and GHs for patients diagnosed with cataract. DATA COLLECTION: Reimbursement claims were extracted from existing claims databases of the largest Dutch health insurer. Quality indicators were obtained by external agencies through a mixed-mode survey. PRINCIPAL FINDINGS: There are no stark differences in complexity of cases for cataract care. ITCs seem to perform surgeries more frequently per care pathway, but conduct a lower number of health care activities per surgical claim. Total average costs are lower in ITCs compared with GHs, but when adjusted for case mix, the differences in costs are lower. The findings with the adjusted quality differences suggest that ITCs outperform GHs on patient satisfaction, but patients' outcomes are similar. CONCLUSION: This finding supports the postulation-based on the focus factory theory-that ITCs can provide more value for cataract care than GHs.

Enabling Analytics on Sensitive Medical Data with Secure Multi-Party Computation.

While there is a clear need to apply data analytics in the healthcare sector, this is often difficult because it requires combining sensitive data from multiple data sources. In this paper, we show how the cryptographic technique of secure multi-party computation can enable such data analytics by performing analytics without the need to share the underlying data. We discuss the issue of compliance to European privacy legislation; report on three pilots bringing these techniques closer to practice; and discuss the main challenges ahead to make fully privacy-preserving data analytics in the medical sector commonplace.