A Privacy-Preserving Infrastructure for Analyzing Personal Health Data in a Vertically Partitioned Scenario

It is widely anticipated that the use and analysis of health-related big data will enable further understanding and improvements in human health and wellbeing. Here, we propose an innovative infrastructure, which supports secure and privacy-preserving analysis of personal health data from multiple providers with different governance policies. Our objective is to use this infrastructure to explore the relation between Type 2 Diabetes Mellitus status and healthcare costs. Our approach involves the use of distributed machine learning to analyze vertically partitioned data from the Maastricht Study, a prospective population-based cohort study, and data from the official statistics agency of the Netherlands, Statistics Netherlands (Centraal Bureau voor de Statistiek; CBS). This project seeks an optimal solution accounting for scientific, technical, and ethical/legal challenges. We describe these challenges, our progress towards addressing them in a practical use case, and a simulation experiment.

Using the Personal Health Train for Automated and Privacy-Preserving Analytics on Vertically Partitioned Data

Conventional data mining algorithms are unable to satisfy the current requirements on analyzing big data in some fields such as medicine, policy making, judicial, and tax records. However, applying diverse datasets from different institutes (both healthcare and non-healthcare related) can enrich information and insights. So far, analyzing this data in an automated, privacy-preserving manner does not exist to our knowledge. In this work, we propose an infrastructure, and proof-of-concept for privacy-preserving analytics on vertically partitioned data.

Acceptance and long-term compliance of nCPAP in obstructive sleep apnea

In a retrospective cohort study, we evaluated whether improvements in nasal continuous positive airway pressure (nCPAP) technology, particularly the introduction of automatic adjustment of the nCPAP pressure (auto-CPAP), have led to better acceptance and (long-term) compliance in patients with obstructive sleep apnea syndrome (OSAS) as compared to earlier reported data. Questionnaires were sent to 256 patients, who were referred to our clinic for an overnight polysomnography from January 1997 to July 2005 and received nCPAP therapy for OSAS. Of the 256 patients, 24 patients were unavailable for follow-up. Of the remaining 232 patients, 58 patients (25%) had discontinued therapy, while 174 patients (75%) were still using nCPAP after 2 months to 8 years of follow-up. One Hundred and thirty eight (79%) of these 174 patients used nCPAP for at least 4 h/night during >or= 5 nights/week, 82,1% of the conventional nCPAP (fixed pressure CPAP) group (n = 78) and 77,1% of the auto-CPAP group (n = 96). Therefore, including the 58 failures, only 59.5% of patients can be seen as compliant. There were no statistical differences between the fixed pressure CPAP and auto-CPAP users, and between the compliant and non-compliant users according to age, BMI, AHI and Epworth sleepiness scale (ESS). Auto-CPAP patients used significantly more cm H2O. The long-term compliance of nCPAP therapy has have increased only slightly since the introduction of the fixed pressure CPAP 25 years ago, in spite of many efforts to improve it. It seems that a plateau has been reached and that it is unrealistic to aim at a substantially higher compliance rate.

Effectiveness of discontinuing antibiotic treatment after three days versus eight days in mild to moderate-severe community acquired pneumonia: randomised, double blind study

OBJECTIVE

To compare the effectiveness of discontinuing treatment with amoxicillin after three days or eight days in adults admitted to hospital with mild to moderate-severe community acquired pneumonia who substantially improved after an initial three days' treatment.

DESIGN

Randomised, double blind, placebo controlled non-inferiority trial.

SETTING

Nine secondary and tertiary care hospitals in the Netherlands.

PARTICIPANTS

Adults with mild to moderate-severe community acquired pneumonia (pneumonia severity index score < or = 110).

INTERVENTIONS

Patients who had substantially improved after three days' treatment with intravenous amoxicillin were randomly assigned to oral amoxicillin (n = 63) or placebo (n = 56) three times daily for five days.

MAIN OUTCOME MEASURES

The primary outcome measure was the clinical success rate at day 10. Secondary outcome measures were the clinical success rate at day 28, symptom resolution, radiological success rates at days 10 and 28, and adverse events.

RESULTS

Baseline characteristics were comparable, with the exception of symptom severity, which was worse in the three day treatment group. In the three day and eight day treatment groups the clinical success rate at day 10 was 93% for both (difference 0.1%, 95% confidence interval--9% to 10%) and at day 28 was 90% compared with 88% (difference 2.0%,--9% to 15%). Both groups had similar resolution of symptoms. Radiological success rates were 86% compared with 83% at day 10 (difference 3%,--10% to 16%) and 86% compared with 79% at day 28 (difference 6%,--7% to 20%). Six patients (11%) in the placebo group and 13 patients (21%) in the active treatment group reported adverse events (P = 0.1).

CONCLUSIONS

Discontinuing amoxicillin treatment after three days is not inferior to discontinuing it after eight days in adults admitted to hospital with mild to moderate-severe community acquired pneumonia who substantially improved after an initial three days' treatment.