Value of a national administrative database to guide public decisions: From the système national d’information interrégimes de l’Assurance Maladie (SNIIRAM) to the système national des données de santé (SNDS) in France

Long-term Persistence of First-line Biologics for Patients With Psoriasis and Psoriatic Arthritis in the French Health Insurance Database

Importance

Treatment options for psoriasis (PsO) and psoriatic arthritis (PsA) have evolved significantly throughout the era of biologics. Clinical trials are inadequate to assess the relative long-term efficacy of biologics and are often insufficient regarding safety.

Objectives

To assess the long-term persistence of different biologic classes to treat PsO and PsA.

Design, Setting, and Participants

This nationwide cohort study involved the administrative health care database of the French health insurance scheme linked to the hospital discharge database. All adults with PsO and PsA who were new users of biologics (not in the year before the index date) from January 1, 2015, to May 31, 2019, were included and followed up through December 31, 2019. Patients hospitalized for PsA in the PsO cohort and for PsO in the PsA cohort in the year before the index date were excluded. Data were analyzed from June 1 to October 31, 2021.

Main Outcomes and Measures

Persistence was defined as the time from biologic therapy initiation to discontinuation and was estimated using the Kaplan-Meier method. Comparison of persistence by biologic class involved using propensity score-weighted Cox proportional hazards regression models and adjustment on specific systemic nonbiologics (time-dependent variables).

Results

A total of 16 892 patients with PsO were included in the analysis (mean [SD] age, 48.5 [13.8] years; 9152 men [54.2%] men). Of these, 10 199 patients (60.4%) started therapy with a tumor necrosis factor (TNF) inhibitor; 3982 (23.6%), with an interleukin 12 and interleukin 23 (IL-12/23) inhibitor; and 2711 (16.0%), with an interleukin 17 (IL-17) inhibitor. An additional 6531 patients with PsA (mean [SD] age, 49.1 [12.8] years; 3565 [54.6%] women) were included; of these, 4974 (76.2%) started therapy with a TNF inhibitor; 803 (12.3%), with an IL-12/23 inhibitor; and 754 (11.5%), with an IL-17 inhibitor. Overall 3-year persistence rates were 40.9% and 36.2% for PsO and PsA, respectively. After inverse probability of treatment weighting and adjustment, the IL-17 inhibitor was associated with higher persistence compared with the TNF inhibitor for PsO (weighted hazard ratio [HR], 0.78 [95% CI, 0.73-0.83]) and PsA (weighted HR, 0.70 [95% CI, 0.58-0.85]) and compared with the IL-12/23 inhibitor for PsA (weighted HR, 0.69 [95% CI, 0.55-0.87]). No difference between the IL-17 inhibitor and IL-12/23 inhibitor for PsO was noted. The IL-12/23 inhibitor was associated with higher persistence than the TNF inhibitor for PsO (weighted HR, 0.76 [95% CI, 0.72-0.80]), with no difference observed for PsA.

Conclusions and Relevance

The findings of this cohort study suggest that IL-17 inhibitors are associated with higher treatment persistence than the TNF inhibitor for PsO and PsA. Interleukin 17 inhibitors were also associated with higher persistence than the IL-12/23 inhibitor for PsA, with no difference for PsO. However, the persistence rates of all biologics remained globally low at 3 years.

Nationwide Initiation of Cardiovascular Risk Treatments During the COVID-19 Pandemic in France: Women on a Slippery Slope?

Objectives

This study examines the initiation of prescribed medication treatments for cardiovascular risk (antihypertensives, lipid-lowering drugs, oral anticoagulants in atrial fibrillation, and smoking cessation medications) during the COVID-19 pandemic in the French population.

Methods

For each year between 2017 and 2021, we used the French National Insurance Database to identify the number of people with at least one reimbursement for these medications but no reimbursement in the previous 12 months. We computed incidence rate ratios (IRRs) between 2017–2019 and, respectively 2020 and 2021 using Poisson regression adjusted for age and 2017–2019 time trends. We recorded the number of lipid profile blood tests, Holter electrocardiograms, and consultations with family physicians or cardiologists.

Results

In 2020, IRR significantly decreased for initiations of antihypertensives (−11.1%[CI95%, −11.4%;−10.8%]), lipid-lowering drugs (−5.2%[CI95%, −5.5%;−4.8%]), oral anticoagulants in atrial fibrillation (−8.6%[CI95%, −9.1%;−8.0%]), and smoking cessation medications (−50.9%[CI95%, −51.1%;−50.7%]) compared to 2017–2019. Larger decreases were found in women compared to men except for smoking cessation medications, with the sex difference increasing with age. Similar analyses comparing 2021 to 2017–2019 showed an increase in the initiation of lipid-lowering drugs (+ 11.6%[CI95%, 10.7%;12.5%]) but even lower rates for the other medications, particularly in women. In addition, the 2020 number of people visiting a family physician or cardiologist decreased by 8.4 and 7.4%. A higher decrease in these visits was observed in those over 65 years of age compared to those under 65 years of age. A greater use of teleconsultation was found in women.

Conclusion

The COVID-19 pandemic heavily impacted the initiation of medication treatments for cardiovascular risk in France, particularly in women and people over 65 years.

The economic burden of infertility treatment and distribution of expenditures overtime in France: a self-controlled pre-post study

Background

Recent cost studies related to infertility treatment have focused on assisted reproductive technologies. None has examined lower-intensity infertility treatments or analyzed the distribution of infertility treatment expenditures over time.

The Purpose of the study was to analyse the size and distribution of infertility treatment expenditures over time, and estimate the economic burden of infertility treatment per 10,000 women aged 18 − 50 in France from a societal perspective.

Methods

We used French National individual medico-administrative database to conduct a self-controlled before-after analytic cohort analysis with 556 incidental women treated for infertility in 2014 matched with 9,903 controls using the exact matching method. Infertility-associated expenditures per woman and per 10,000 women over the 3.5-year follow-up period derived as a difference-in-differences.

Results

The average infertility related expenditure per woman is estimated at 6,996 (95% CI: 5,755–8,237) euros, the economic burden for 10,000 women at 70.0 million (IC95%: 57.6–82.4) euros. The infertility related expenditures increased from 235 (IC95%: 98–373) euros in semester 0, i.e. before treatment, to 1,509 (IC95%: 1,277–1,741) euros in semester 1, mainly due to ovulation stimulation treatment (47% of expenditure), to reach a plateau in semesters 2 (1,416 (IC95%: 1,161–1,670)) and 3 (1,319 (IC95%: 943–1,694)), where the share of expenses is mainly related to hospitalizations for assisted reproductive technologies (44% of expenditure), and then decrease until semester 6 (577 (IC95%: 316–839) euros).

Conclusion

This study informs public policy about the economic burden of infertility estimated at 70.0 million (IC95%: 57.6–82.4) euros for 10,000 women aged between 18 and 50. It also highlights the importance of the share of drugs in infertility treatment expenditures. If nothing is done, the increasing use of infertility treatment will lead to increased expenditure. Prevention campaigns against the preventable causes of infertility should be promoted to limit the use of infertility treatments and related costs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-022-07725-9.

Analysis of vitreoretinal surgery activity in metropolitan France in 2016: impact on training capacities

PURPOSE

The purpose of the study was to describe vitreoretinal surgery activity and vitreoretinal surgeons in private or public practice in metropolitan France over the year 2016 to anticipate surgical training needs.

METHODS

Patients aged ≥20 years who had undergone vitreoretinal surgery, alone or combined with cataract surgery were included using the French National Healthcare system database. For surgery performed by ophthalmologists carrying out ≥50 procedures during the year, the incidence per 100 000 of population ≥ 20 years of age, the number and mean age of surgeons and the number of surgeons aged >55 years were calculated.

RESULTS

Overall, 57 947 posterior segment surgical procedures were included, 40% in the public sector and 49% in the private sector for private surgeons and/or public centres performing ≥50 procedures/year. The remaining 11% of procedures were from private surgeons and/or public centres performing <50 procedures/year. The analysis included 356 surgeons with a mean age of 41 ± 10 years (39% female) in the public sector and 47 ± 10 years (14% female) in the private sector. The majority of urgent surgery was for retinal detachment (n = 30 290 [52% of total surgical procedures]). Scheduled surgery involved surgery for macular holes and epiretinal membranes (n = 16 454 [28% of total surgical procedures]). Combined vitrectomy-phacoemulsification surgery (n = 10 120) represented 17% of all vitreoretinal surgery. University regions with the fewest surgeons and regions with surgeons >55 years of age were identified, to anticipate the training need for new surgeons.

CONCLUSION

This study demonstrated disparities in the geographic distribution of vitreoretinal surgery in France and identified regions that need increased training capacities to ensure a sufficient number of surgeons.

Healthcare cost by primary tumour, functioning status and treatment among patients with metastatic neuroendocrine tumours: The LyREMeNET study

The annual prevalence of metastatic neuroendocrine tumours (mNETs) is rising, leading to significant healthcare costs. The present study aimed to describe healthcare resource use (HRU) and the corresponding costs among patients with mNETs, according to primary tumour location, functioning status and type of treatments. The LyREMeNET study included consecutive mNET patients with a diagnosis performed between January 2010 and December 2017, who were seen at least once in the ENETS center of excellence in Lyon. The median HRU and costs per patient were estimated, up to 3 years before and after the diagnosis. The Cancer database of the center was linked to the French national health data system. HRU and related costs were described per person per month (PPPM). Among 316 patients presenting with a mNET, 48.4% had a small-intestinal mNET, 32.3% had a pancreatic mNET and 39.2% had carcinoid syndrome. The mean overall cost increased from €615 to €2875 PPPM between the years preceding and following the diagnosis, and remained above €2500 in the two subsequent years. The two main cost drivers of total healthcare expenditure were drugs (€1161) and hospital stay (€662). Median costs of mNETs arising from pancreas and small intestine were €2325 and €2540 PPPM, respectively. Costs were higher in patients with a functional mNET (€2807 PPPM for carcinoid syndrome) and during peptide receptor radionuclide therapy (PRRT) (€8835 PPPM). The highest overall cost was found during the first year following the diagnosis. Cost of care was higher for small intestine mNETs, for functional mNETs and during peptide receptor radionuclide therapy.

Functional Outcome After Spinal Meningioma Surgery. A Nationwide Population-Based Study

Objective

To describe and analysed the functional outcome (FO) after spinal meningioma (SM) surgery.

Methods

We processed the système national des données de santé (SNDS) i.e. , the French national administrative medical database to retrieve appropriate cases. We analysed the International Classification of Diseases 10 codes to assess the FO. Logistic models were implemented to search for variables associated with a favourable FO i.e. , a patient being independent at home without disabling symptom.

Results

A total of 2,844 patients were identified of which 79.1% were female. Median age at surgery was 66 years, interquartile range (IQR) (56–75). Ninety-five point nine percent of the SMs were removed through a posterior ± lateral approach and 0.7% need an associated stabilisation. Benign meningioma represented 92.9% and malignant 2.1%. Median follow-up was 5.5 years, IQR (2.1–8), and at data collection 9% had died. The FO was good and increased along the follow-up: 84.3% of the patients were alive and had not associated symptoms at one year, 85.9% at 2 and 86.8% at 3 years. Nonetheless, 3 years after the surgery 9.8% of the alive patients still presented at least one disabling symptom of which 2.7% motor deficit, 3.3% bladder control problem, and 2.5% gait disturbance. One point seven percent were care-provider dependent and 2.1% chair or bedfast. In the multivariable logistic regression an older age at surgery (odds ratio [OR], 0.37; 95% confidence interval [CI], 0.29–0.47, p<0.001), a high level of comorbidities (OR, 0.71; 95% CI, 0.66–0.75, p<0.001), and an aggressive tumor (OR, 0.49; 95% CI, 0.33–0.73; p<0.001) were associated with a worse FO.

Conclusion

FO after meningioma surgery is favourable but, may be impaired for older patients with a high level of comorbidities and aggressive tumor.

Impact of a school-based and primary care-based multicomponent intervention on HPV vaccination coverage among French adolescents: a cluster randomised controlled trial protocol (the PrevHPV study)

INTRODUCTION

Vaccination is an effective and safe strategy to prevent Human papillomavirus (HPV) infection and related harms. Despite various efforts by French authorities to improve HPV vaccine coverage (VC) these past few years, VC has remained far lower than in most other high-income countries. To improve it, we have coconstructed with stakeholders a school-based and primary care-based multicomponent intervention, and plan to evaluate its effectiveness, efficiency and implementation through a cluster randomised controlled trial (cRCT).

METHODS AND ANALYSIS

This pragmatic cRCT uses an incomplete factorial design to evaluate three components applied alone or in combination: (1) adolescents and parents' education and motivation at school, using eHealth tools and participatory learning; (2) general practitioners' training on HPV using motivational interviewing techniques and provision of a decision aid tool; (3) free-of-charge access to vaccination at school. Eligible municipalities (clusters) are located in one of 14 preselected French school districts and must have only one secondary school which enrols at least 2/3 of inhabitants aged 11-14 years. A randomisation stratified by school district and deprivation index allocated 90 municipalities into 6 groups of 15. The expected overall sample size estimate is 41 940 adolescents aged 11-14 years. The primary endpoint is the HPV VC (≥1 dose) among adolescents aged 11-14 years, at 2 months, at the municipality level (data from routine databases). Secondary endpoints include: HPV VC (≥1 dose at 6 and 12 months; and 2 doses at 2, 6 and 12 months); differences in knowledge, attitudes, behaviours, and intention among adolescents, parents and general practitioners between baseline and 2 months after intervention (self-administered questionnaires); incremental cost-effectiveness ratio. Implementation measures include dose, fidelity, adaptations, reached population and satisfaction (activity reports and self-administered questionnaires).

ETHICS AND DISSEMINATION

This protocol was approved by the French Ethics Committee 'CPP Sud-Est VI' on 22 December 2020 (ID-RCB: 2020-A02031-38). No individual consent was required for this type of research; all participants were informed of their rights, in particular not to participate or to oppose the collection of data concerning them. Findings will be widely disseminated (conference presentations, reports, factsheets and academic publications).

TRIAL REGISTRATION NUMBER

NCT04945655.

Risk of first ischaemic stroke and use of antidopaminergic antiemetics: nationwide case-time-control study

OBJECTIVE

To estimate the risk of ischaemic stroke associated with antidopaminergic antiemetic (ADA) use.

DESIGN

Case-time-control study.

SETTING

Data from the nationwide French reimbursement healthcare system database Système National des Données de Santé (SNDS).

PARTICIPANTS

Eligible participants were ≥18 years with a first ischaemic stroke between 2012 and 2016 and at least one reimbursement for any ADA in the 70 days before stroke. Frequencies of ADA reimbursements were compared for a risk period (days -14 to -1 before stroke) and three matched reference periods (days -70 to -57, -56 to -43, and -42 to -29) for each patient. Time trend of ADA use was controlled by using a control group of 21 859 randomly selected people free of the event who were individually matched to patients with stroke according to age, sex, and risk factors of ischaemic stroke.

MAIN OUTCOME MEASURES

Association between ADA use and risk of ischaemic stroke was assessed by estimating the ratio of the odds ratios of exposure evaluated in patients with stroke and in controls. Analyses were adjusted for time varying confounders (anticoagulants, antiplatelets, and prothrombotic or vasoconstrictive drugs).

RESULTS

Among the 2612 patients identified with incident stroke, 1250 received an ADA in the risk period and 1060 in the reference periods. The comparison with the 5128 and 13 165 controls who received an ADA in the same periods yielded a ratio of adjusted odds ratios of 3.12 (95% confidence interval 2.85 to 3.42). Analyses stratified by age, sex, and history of dementia showed similar results. Ratio of adjusted odds ratios for analyses stratified by ADA was 2.51 (2.18 to 2.88) for domperidone, 3.62 (3.11 to 4.23) for metopimazine, and 3.53 (2.62 to 4.76) for metoclopramide. Sensitivity analyses suggested the risk would be higher in the first days of use.

CONCLUSIONS

Using French nationwide exhaustive reimbursement data, this self-controlled study reported an increased risk of ischaemic stroke with recent ADA use. The highest increase was found for metopimazine and metoclopramide.

Patients Hospitalized for Ischemic Stroke and Intracerebral Hemorrhage in France: Time Trends (2008-2019), In-Hospital Outcomes, Age and Sex Differences

Background: Rates of patients hospitalized for stroke increased among people aged under 65 years in France, as has been found in other countries. Methods: To analyze time trends in the rates of patients hospitalized for ischemic stroke (IS) and intracerebral hemorrhage (ICH) in France between 2008 and 2019 and determine related short-term outcomes mainly, we selected all patients hospitalized for stroke using the French national hospital database. Results: The average annual percentage change in the rates of patients hospitalized for IS increased significantly in men and women aged 50–64 years (+2.0%) and in men aged 18–34 years (+1.5%) and 35–44 years (+2.2%). A decrease in the average annual percentage change was observed for IS among people aged over 75 years and among those over 50 years for ICH. After adjustment on confounding factors, women were less likely to die in hospital. Case fatality rates decreased overtime in all age groups for both sexes, with a more pronounced decrease for IS than ICH. Conclusions: The increasing trend of IS among adults under 65 years is ongoing, highlighting the urgent need for stroke prevention programs in that age. For the first time, we recorded a decrease in the rates of patients hospitalized for ICH among the population over 50 years.

One Year Prevalence of Venous Thromboembolism in Hospitalized COVID-19 Patients in France: Patients' Characteristics, Time Trends, and Outcomes

BACKGROUND

 Patients hospitalized with coronavirus disease-2019 (COVID-19) are at high risk of deep venous thrombosis (DVT) and pulmonary embolism (PE).

OBJECTIVES

 The aims were to provide time trends in the 2020 nation-wide prevalence of venous thromboembolism (VTE) in patients hospitalized with a COVID-19 diagnosis in France, and to describe in-hospital and up to 30-day postdischarge death.

METHODS

 All patients hospitalized in France with a COVID-19 diagnosis in 2020 were selected. Crude and age-adjusted prevalence of VTE and PE was computed by 4-week intervals and for the overall study period using Poisson regression. Time trends in in-hospital and 30-day postdischarge case-fatality rates were evaluated by comparing each 4-week intervals to weeks 10 to 14 corresponding to the first part of the first lockdown using logistic regression models.

RESULTS

 Among the 287,638 patients hospitalized with a COVID-19 diagnosis in 2020 in France, 14,985 (5.2%) had a concomitant VTE, with 10,453 (3.6%) having PE and 4,532 (1.6%) having DVT. In patients admitted to intensive care units, the crude prevalence of VTE and PE reached 16.1 and 11.0% respectively during the first lockdown. After adjustment, the prevalence of VTE and PE decreased during the year 2020 but a rebound was observed during the second lockdown. In-hospital case-fatality rates among hospitalized COVID-19 patients with PE globally decreased between the first and the second epidemic waves.

CONCLUSION

 Our study showed a decrease in the incidence of symptomatic VTE and PE in hospitalized COVID-19 patients, and a decreased time trend of outcomes during the second wave compared with the first one.

Monitoring the impact of COVID-19 in France on cancer care: a differentiated impact

The COVID-19 pandemic has had a substantial and lasting impact on care provision, particularly in the field of cancer care. National steering has helped monitor the health situation and adapt the provision and organisation of care. Based on data from the French administrative healthcare database (SNDS) on the entire French population (67 million people), screening, diagnostic and therapeutic activity was monitored and compared 2019 on a monthly basis. A noteworthy decline in all activities (with the exception of chemotherapy) was observed during the first lockdown in France. Over the months that followed, this activity returned to normal but did not make up for the shortfall from the first lockdown. Finally, during the lockdown in late 2020, cancer care activity was conserved. In brief, in 2020, the number of mammograms decreased by 10% (− 492,500 procedures), digestive endoscopies by 19% (− 648,500), and cancer-related excision by 6% (− 23,000 surgical procedures). Hospital radiotherapy activity was down 3.8% (− 4400 patients) and that in private practice was down 1.4% (− 1600 patients). Chemotherapy activity increased by 2.2% (7200 patients), however. To summarize, COVID-19 had a very substantial impact during the first lockdown. Safeguarding cancer care activity helped limit this impact over the months that followed, but the situation remains uncertain. Further studies on the medium- and long-term impact on individuals (survival, recurrence, after-effects) will be conducted.

Exposure to High Precariousness Prevalence Negatively Impacts Drug Prescriptions of General Practitioners to Precarious and Non-Precarious Populations: A Retrospective Pharmaco-Epidemiological Study

(1) Background: Precarious patients are more difficult to care for due to low literacy rates and poor adherence to treatment and hospitalization. These difficulties have detrimental effects on general practitioners (GPs), deteriorating medical communication, advice, diagnoses, and drug prescriptions. To better understand how precariousness affects primary care, we tested whether, among GPs, exposure to high precariousness prevalence more severely impacts drug prescriptions to precarious and non-precarious populations compared to low precariousness prevalence. Materials and methods: This pharmaco-epidemiological study, using linear regression analyses, compared the defined daily dose of 20 drugs prescribed by GPs to precarious and non-precarious patients in four French regions with low and high precariousness prevalence in 2015. (2) Findings: Exposure to high precariousness prevalence significantly impacted the prescriptions of nine medications to precarious patients and two medications to non-precarious patients, and distributed into three interaction patterns. (3) Interpretation: The selective over-prescription of drugs with easy intake modalities to precarious patients probably reflects GPs’ attempts to compensate for poor patient compliance. In contrast, the under-prescription of drugs targeting fungal infections in precarious populations and diabetes and cardiovascular diseases in non-precarious populations was seemingly due to a breakdown of empathy and professional exhaustion, causing medical neglect.

Potential determinants of unfavourable healthcare utilisation trajectories during the last year of life of people with incident Alzheimer Disease or Related Syndromes: a nationwide cohort study using administrative data

BACKGROUND

people approaching the end-of-life frequently face inappropriate care. With Alzheimer Disease or Related Syndromes (ADRS), end-of-life is characterised by progressive decline, but this period remains difficult to identify. This leads to a lack of anticipation and sometimes with unfavourable healthcare utilisation trajectories (HUTs).

OBJECTIVE

to quantify unfavourable HUTs during the last year of life and identify their potential determinants in both community and nursing-home settings.

DESIGN

nationwide cohort study using administrative database.

SETTING

French community and nursing-home residents.

SUBJECTS

incident ADRS people identified in 2012, who died up to 31 December 2017.

METHODS

we used multidimensional clustering to identify 15 clusters of HUTs, using 11 longitudinal healthcare dimensions during the last year of life. Clusters were qualitatively assessed by pluri-disciplinary experts as favourable or unfavourable HUTs. Individual and contextual potential determinants of unfavourable HUTs were studied by setting using logistic random-effect regression models.

RESULTS

62,243 individuals died before 31 December 2017; 46.8% faced unfavourable end-of-life HUTs: 55.2% in the community and 31.8% in nursing-homes. Individual potential determinants were identified: younger age, male gender, ADRS identification through hospitalisation, shorter survival, life-limiting comorbidities, psychiatric disorders, acute hospitalisations and polypharmacy. In the community, deprivation and autonomy were identified as potential determinants. Contextual potential determinants raised mostly in the community, such as low nurse or physiotherapist accessibilities.

CONCLUSIONS

Nearly half of people with ADRS faced unfavourable HUTs during their last year of life. Individual potential determinants should help anticipate advance care planning and palliative care needs assessment. Contextual potential determinants suggest geographical disparities and health inequalities.

Measuring severe neonatal morbidity using hospital discharge data in France

BACKGROUND

Measuring infant health at birth is key for surveillance and research in obstetrics and neonatology, but there is no international consensus on morbidity indicators. The Neonatal Adverse Outcome Indicator (NAOI) is a composite indicator, developed in Australia, which measures the burden of severe neonatal morbidity using hospital discharge data.

OBJECTIVE

To evaluate the applicability of the NAOI in France for surveillance and research.

METHODS

We constituted a cohort of live births ≥24 weeks' gestational age in Metropolitan France from 2014 to 2015 using hospital discharge, insurance claims and cause of death data. Outlier hospitals were identified using funnel plots of standardised morbidity ratios (SMR), and their coding patterns were assessed. We compared the NAOI and its component codes with published Australian and English data and estimated unadjusted and adjusted risk ratios for known risk factors for neonatal morbidity.

RESULTS

We included 1,459,123 births (511 hospitals). Twenty-eight hospitals had SMR above funnel plot control limits. Newborns with NAOI morbidities in these hospitals had lower mortality and shorter stays than in other hospitals. Amongst within-limit hospitals, NAOI prevalence was 4.8%, comparable to Australia (4.6%) and England (5.4%). Most individual components had a similar prevalence, with the exception of respiratory support, intravenous fluid procedures and infection. NAOI was lowest at 39 weeks (2.2%) with higher risks for maternal age ≥40 (relative risk [RR] 1.47, 95% confidence interval [CI] 1.42, 1.51), state medical insurance (RR 1.60, 95% CI 1.52, 1.68), male sex (RR 1.21, 95% CI 1.19, 1.23) and birthweight <3^rd percentile (RR 4.60, 95% CI 4.51, 4.69).

CONCLUSIONS

The NAOI provides valuable information on population prevalence of severe neonatal morbidity and its risk factors. Whilst the prevalence was similar in high-income countries with comparable neonatal mortality levels, ensuring valid comparisons between countries and hospitals will require further work to harmonize coding procedures, especially for infection and respiratory morbidity.

Estimating the Future Burden of Myocardial Infarction in France Until 2035: An Illness-Death Model-Based Approach

Purpose

In France, myocardial infarction (MI) was the second leading cause of years of life lost in 2019. Estimating the burden of MI in future years could help policymakers and other actors anticipate care and prevention needs and guide them in public health decision-making.

Materials and Methods

Using data from the French hospital discharge database from 2007 to 2015 (n = 519,400), demographic data, and an illness-death model, we projected incidence, prevalence, number of prevalent cases and mean age of incident MI cases in France. The methodology took into account the age-cohort effect on MI incidence, mortality of healthy and diseased subjects, and the time since disease onset.

Results

Projections highlighted an increase in MI prevalence in men between 2015 and 2035 from 2.52% (95% uncertainty interval (UI): [2.48–2.56]) in 2015 to 4.02% ([3.92–4.12]) in 2035, and from 0.85% ([0.83–0.87]) to 1.44% ([1.38–1.50]) in women. This corresponds to an increase of 365,000 cases between 2015 and 2035 (+81.1%) for men and 146,000 cases for women (+88.0%). The difference in the mean age of incident cases between men and women decreased from 9.52 in 2015 to 5.49 years in 2035.

Conclusion

Our projections forecast an increase in MI prevalence between 2015 and 2035 in men and women, especially in relatively younger women. Using statistical models such as ours can help assess the impact of prevention campaigns for the main cardiovascular disease risk factors on the future MI prevalence.

Machine learning versus regression modelling in predicting individual healthcare costs from a representative sample of the nationwide claims database in France

BACKGROUND

Innovative provider payment methods that avoid adverse selection and reward performance require accurate prediction of healthcare costs based on individual risk adjustment. Our objective was to compare the performances of a simple neural network (NN) and random forest (RF) to a generalized linear model (GLM) for the prediction of medical cost at the individual level.

METHODS

A 1/97 representative sample of the French National Health Data Information System was used. Predictors selected were: demographic information; pre-existing conditions, Charlson comorbidity index; healthcare service use and costs. Predictive performances of each model were compared through individual-level (adjusted R-squared (adj-R²), mean absolute error (MAE) and hit ratio (HiR)), and distribution-level metrics on different sets of covariates in the general population and by pre-existing morbid condition, using a quasi-Monte Carlo design.

RESULTS

We included 510,182 subjects alive on 31st December, 2015. Mean annual costs were 1894€ (standard deviation 9326€) (median 393€, IQ range 95€; 1480€), including zero-claim subjects. All models performed similarly after adjustment on demographics. RF model had better performances on other sets of covariates (pre-existing conditions, resource counts and past year costs). On full model, RF reached an adj-R² of 47.5%, a MAE of 1338€ and a HiR of 67%, while GLM and NN had an adj-R² of 34.7% and 31.6%, a MAE of 1635€ and 1660€, and a HiR of 58% and 55 M, respectively. RF model outperformed GLM and NN for most conditions and for high-cost subjects.

CONCLUSIONS

RF should be preferred when the objective is to best predict medical costs. When the objective is to understand the contribution of predictors, GLM was well suited with demographics, conditions and base year cost.

Incidence and Prevalence of Neovascular Age-Related Macular Degeneration in France between 2008 and 2018

Purpose

This study aimed to estimate the incidence and prevalence of neovascular age-related macular degeneration (nAMD) in the French population between 2008 and 2018.

Design

This was a retrospective, longitudinal population study using health care consumption data from the Système National des Données de Santé (SNDS; the French National Health Information Database), which covers approximately 99% of the French population.

Participants

We identified individuals treated for nAMD from the French population 50 years of age and older. Identification criteria were nAMD diagnosis or reimbursement of nAMD treatments (anti–vascular endothelial growth factor intravitreal injection or dynamic phototherapy with verteporfin). Exclusion criteria were high myopia, diagnosis of other retinal diseases, and other treatments for macular diseases (dexamethasone implant, laser therapy, etc.).

Methods

We calculated incidence and prevalence based on the age-matched general population in France. Adjustment for age and sex was also performed for incidence.

Main Outcome Measures

Incidence and prevalence of nAMD in the French population between 2008 and 2018.

Results

Between 2008 and 2018, we identified 342 961 patients with nAMD (67.5% women). Mean ± standard deviation age at nAMD diagnosis or first treatment increased from 78.8 ± 8.1 years in 2008 to 81.2 ± 7.9 years in 2018. In 2018, annual incidence was 0.149% and prevalence was 1.062% for the French population 50 years of age or older. Incidence was stable over the 10-year period. Annual incidence increased with age (0.223%, 0.380%, and 0.603% in those 60 years of age or older, 70 years of age or older, and 80 years of age or older, respectively), with similar trends for prevalence. No major differences were observed among the 14 regions of France for incidence or prevalence. Neovascular age-related macular degeneration incidence in 2018 was not impacted by the availability of primary or ophthalmology care in patients’ localities.

Conclusions

The LANDSCAPE study provides exhaustive nationwide data on incidence and prevalence of nAMD in France over a 10-year period.

National Health Care data system analysis of glaucoma surgery activity in France in 2016

PURPOSE

To describe the glaucoma surgery offer in France in 2016.

METHODS

We used the French National Health Care System database to identify all medical procedures carried out in 2016. The study investigated the entire population aged 30 years and older that had undergone glaucoma surgery, alone or combined with another surgery. We calculated the incidence of surgeries per 100 000 inhabitants 30 years of age and older performed by ophthalmologists carrying out at least 50 procedures annually, the number of surgeons doing these surgeries, the mean age of these practitioners, and the number of surgeons older than 55 years.

RESULTS

In 2016, 16 854 glaucoma surgeries were performed in patients aged 30 years and older, for an incidence of 40.8 per 100 000 inhabitants aged 30 years and older. The most frequent procedure performed was trabeculectomy followed by non-penetrating deep sclerectomy (16.7 and 11.7, respectively, per 100 000 inhabitants 30 years of age and older). Private practice glaucoma surgery accounted for 47% of the activity of surgeons performing at least 50 surgeries per year and 60% of the total surgical activity. Of the private practice ophthalmologists performing at least 50 glaucoma surgery procedures per year, 58.5% were over 55 years of age, and 23.5% of public hospital ophthalmologists were over 55 years of age.

CONCLUSIONS

This study demonstrates that surgeons performing glaucoma surgeries are often older. It is necessary to take note of the country's educational capacity to ensure that the number of ophthalmological surgeons remains adapted to demand.

Effectiveness of a multi-faceted intervention to deprescribe proton pump inhibitors in primary care: protocol for a population-based, pragmatic, cluster-randomized controlled trial

Background

Inappropriately using proton pump inhibitors (PPI) is associated with severe adverse drug reactions and may have major consequences on healthcare costs. Deprescribing (the process by which a healthcare professional supervises the withdrawal of an inappropriate medication, to manage polypharmacy and improve outcomes) should be considered when an inappropriate PPI prescription is identified. Deprescribing interventions directed solely to prescribers have limited efficacy and are rarely targeted to patients. The aim of this trial is to assess the efficacy of a multi-faceted intervention with patients and general practitioners (GPs) to deprescribe PPI.

Methods

We will conduct a pragmatic, cluster-randomized, population-based, controlled trial in two regions of Western France. GPs with practices with over 100 patients, and their adult patient to whom over 300 defined daily doses (DDD) of PPIs have been dispensed in the year before baseline will be included. A total of 1300 GPs and 33,000 patients will be cluster-randomized by GPs practices. Three arms will be compared: i) a multi-faceted intervention associating a) a patient education brochure about PPI deprescribing sent directly to patients (the brochure was designed using a mixed-methods study), and b) a personalized letter with the Bruyere’s PPI deprescribing algorithm sent to their respective GPs, or ii) a single intervention where only the GPs received the letter and algorithm, or iii) no intervention.

The primary outcome will be PPI deprescribing, defined as the proportion of patients achieving at least a 50% decrease in the amount of PPI dispensed to them (DDD/year) at 12 months compared to baseline. Secondary outcomes will include incremental cost-utility ratio (using EQ-5D-5L scale and National Health Insurance’s database), acid rebound (using the Gastroesophageal Reflux Disease Impact Scale), and the patients’ attitudes towards deprescribing (using the French rPATD).

Discussion

Based on previous trials, we anticipate more than 10% “successful PPI deprescribing” in the multi-faceted intervention compared to the single intervention on GPs and the control arm. The study has been funded through a national grant and will be launched in autumn 2020, for early results by the end of 2022.

Trial registration

Clinicaltrials.gov NCT04255823; first registered on February 5, 2020.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-022-07496-3.

Self-reported functional assessment after treatment for prostate cancer: 5-year results of the prospective cohort VICAN

Objective: We aimed to assess the long-term association of therapeutic strategies with urinary, sexual function and health-related quality of life (HR-QOL) for 5-year prostate cancer (PC) survivors. Materials & methods: The VICAN survey consisted of self-reported data prospectively collected, including living conditions, treatment side effects and quality of life (QOL) of cancer survivors. Results: Among the 434 PC survivors, 52.8% reported urinary incontinence (UI) and 55.8% reported erectile dysfunction (ED). Patients treated with radical prostatectomy with salvage radiotherapy reported significantly more UI (p = 0.014) and more ED (p = 0.012) compared with other strategies. UI was significantly associated with physical and mental health-related QOL (p = 0.045 and p = 0.049, respectively). Conclusion: Self-assessed functional outcomes 5 years after PC diagnosis remain poor and could have an impact on health-related QOL.

The Evolution of HIV Patient Retention and Care in French Guiana: A Broader View From the Système National des Données de Santé

Background

Although the simplification of antiretroviral (AVR) treatment regimens and follow-up has led to fewer constraints for patients with HIV, their follow-up remains of paramount importance to optimize AVR therapy, to detect and prevent HIV-related morbidity, and prevent secondary infections. The problem of follow-up interruption in French Guiana has been persistent and seemingly impervious to efforts to alleviate it.

Objective

The objective was to follow the trend of follow-up interruptions and to test the hypothesis that an increasing number of patients was, in fact, followed by private practitioners.

Method

Using the complementary lenses of the hospital HIV cohort and the health insurance information system, we looked at the incidence of follow-up interruption and the proportion of patients followed by private practitioners.

Results

We tallied 803 persons that were not known to have died and who were lost to follow-up. Over time, hospital outpatients were lost to follow-up significantly sooner. By contrast, there was a significant trend with more and more patients exclusively followed by private practitioners.

Conclusion

While hospital outpatient care remains by far the most common mode of patient care, there seems to be a gradual erosion of this model in favor of private practice.

Machine Learning-Based Analysis of Treatment Sequences Typology in Advanced Non-Small-Cell Lung Cancer Long-Term Survivors Treated With Nivolumab

PURPOSE

Immune checkpoint inhibitors substantially changed advanced non-small-cell lung cancer (aNSCLC) management and can lead to long-term survival. The aims of this study were (1) to use a machine learning method to establish a typology of treatment sequences on patients with aNSCLC who were alive 2 years after initiating a treatment with anti-programmed death-ligand 1 monoclonal antibody nivolumab and (2) to describe the patients' characteristics according to the typology of treatment sequences.

MATERIALS AND METHODS

This retrospective observational study was based on data from the comprehensive French hospital discharge database for all patients with lung cancer with at least one line of platinum-based chemotherapy, starting nivolumab between January 1, 2015, and December 31, 2016, and alive 2 years after nivolumab treatment initiation. Patients were followed until December 31, 2018. A typology of most common treatment sequences was established using hierarchical clustering with time sequence analysis.

RESULTS

Two thousand two hundred twelve study patients were, on average, 63.0 years old, 69.9% of them were men, and 61.9% had a nonsquamous cell carcinoma. During the 2 years after nivolumab treatment initiation, clusters of patients with four basic types of treatment sequences were identified: (1) almost continuous nivolumab treatment (44% of patients); (2) nivolumab most of the time followed by a treatment-free interval or a chemotherapy (15% of patients); and a short or medium nivolumab treatment, followed by (3) a long systemic treatment-free interval (17% of patients) or (4) a long chemotherapy (23% of patients).

CONCLUSION

This machine learning approach enabled the identification of a typology of four representative treatment sequences observed in long-term survival. It was noted that most long-term survivors were treated with nivolumab for well over 1 year.

Long-term prognosis following vascular graft infection: a 10-year cohort study

Background

Vascular graft infection (VGI) remains a severe disease with high mortality and relapse rates. We performed a retrospective single-center cohort study to highlight factors associated with long-term all-cause mortality in patients with vascular graft infection.

Methods

All patients hospitalized in our facility over 10 years for VGI were included. VGI was defined by the presence of a vascular graft or an aortic stent graft (stent or fabric), associated with 2 criteria among clinical, biological, imaging, or microbiological elements in favor of VGI. The primary outcome was all-cause mortality. Empirical antibiotic therapy was considered as appropriate when all involved pathogens were susceptible in vitro to the antibiotics used. The surgical strategy was defined as nonoptimal when the graft was not removed in a late-onset surgery (>3 months) or no surgery was performed.

Results

One hundred forty-six patients were included. Empirical antibiotic therapy was administered in 98 (67%) patients and considered appropriate in 55 (56%) patients. Surgery was performed in 136 patients (96%) and considered as optimal in 106 (73%) patients. In multivariable analysis, appropriate empirical antibiotic therapy was associated with a lower probability of mortality (hazard ratio, 0.47 [95% confidence interval, .30–.79]; P = .002). Long-term survival did not differ according to whether the surgical strategy was considered optimal or not (log-rank = 0.66).

Conclusions

Appropriate empirical antibiotic therapy is a cornerstone of the management of VGI. Whenever possible, antibiotics must be associated with optimal surgical management. However, surgery could potentially be avoided in comorbid patients who are treated with appropriate antibiotics.

The ongoing French BaMaRa-BNDMR cohort: implementation and deployment of a nationwide information system on rare disease

BACKGROUND

BaMaRa allows the secure collection and deidentified centralization of medical data from all patients followed-up in a rare disease expert network in France, based on a minimum data set (SDM-MR). The present article describes BaMaRa information system implementation and development across the whole national territory as well as data access requests through BNDMR, the data warehouse which centralizes all BaMaRa data, during the 2015-2020 period.

MATERIALS AND METHODS

SDM-MR is made up of 60 interoperable items and is routinely collected through BaMaRa in rare disease centers as part of care and discharged into BNDMR after deidentification and data reconciliation. Data access is regulated by a scientific committee.

RESULTS

In total, 668 002 affected patients had an SDM-MR recorded in BNDMR by the end of 2020 with a mean value of 3.4 activities per patients. Data access was provided for 66 projects.

CONCLUSION

The BaMaRa-BNDMR infrastructure provides an administrative and epidemiological resources for rare diseases in France.

Mortality and Respiratory-Related Hospitalizations in Idiopathic Pulmonary Fibrosis Not Treated With Antifibrotics

Background: Real-world data regarding outcomes of idiopathic pulmonary fibrosis (IPF) are scarce, outside of registries. The claims data from the French National Health System (SNDS) were used to describe outcomes in patients diagnosed with IPF in 2015–2016 but who did not receive antifibrotic therapies.

Method: Patients aged <50 years were excluded, as were patients with pulmonary fibrosis other than IPF, patients who had previously received a lung transplant, and those who had received antifibrotic therapies at any time between 2010 and 2016. Patients were followed-up until their last health record, lung transplantation, initiation of antifibrotic therapies, death, or the end of the study period (31 December 2017), whichever occurred first.

Results: A total of 5,360 patients (43.2%) not treated with antifibrotic therapies were included. The mean age was 75.5 years, and 57.9% were males. In the year before inclusion, 47.3% of patients had a Charlson score ≥5. During follow-up, 41.2% of patients died. The unadjusted incidence rate was 29.9 per 100 person-years (95%CI = [28.7–31.2]), and the cumulative incidence of death at 3 years was 50.2% (95% CI = [48.3–52.1%]). In the study population, 35.3% of patients experienced an acute respiratory-related hospitalization. The unadjusted incidence rate was 32.1 per 100 person-years (95%CI = [30.6–33.5]) and the cumulative incidence of the event at 3 years was 41.5% (95% CI = [39.7–43.2%]).

Interpretation: This observational study showed that, if untreated with antifibrotics, IPF is associated with a 50% all-cause mortality at 3 years. These figures can serve as a historical control of the natural course of the disease.

Risk of pacemaker implantation after radiotherapy for breast cancer: A study based on French nationwide health care database sample

Background

Among cardiac complications of breast cancer radiotherapy (BC RT), there are very limited data on arrhythmia and conduction disorders, in particular severe cases requiring permanent pacemaker implantation (PPMI). Therefore, this exploratory study aimed to evaluate the risk of PPMI for BC patients treated with RT, compared with the general population and with BC patients not treated with RT.

Methods

The study was performed on a 1/97 representative sample of the French health care database (EGB database). Adult women with a first BC treated with or without RT between 2008 and 2016 were included, followed until 2018, and de novo PPMI were identified. We compared the PPMI incidence in BC cohort relative to the general population with standardized incidence ratio (SIR) and evaluated the risk of PPMI in RT patients compared to patients without RT with a competing risk survival analysis.

Results

A total of 3853 BCE patients were included. Among BC patients treated with RT, 28 PPMI cases were observed compared with 13 expected cases, corresponding to a SIR of 2.18 [95% CI: 1.45-3.06]. For BC patients not treated with RT, the SIR was 1.01 [95% CI: 0.40-1.90]. Patients treated with RT showed a borderline significant higher risk of PPMI compared with those not treated with RT (subdistribution Hazard Ratio = 2.08, 95% CI 0.87-4.97, p = 0.09).

Conclusions

Our exploratory findings indicate that, over the last decade in France, BC patients treated with RT appeared to be at higher risk of PPMI than general population. Further studies are needed to expand on this topic.

Cocooning strategy: Pertussis vaccination coverage rate of parents with a new-born in 2016 and 2017 in France

BACKGROUND

The "cocooning" strategy was introduced in 2004 to protect infants too young to be vaccinated against pertussis, by immunizing their parents and close relatives. The study objective was to assess its implementation 12 years after its introduction by estimating the pertussis vaccination coverage rates (VCR) among parents of newborns.

MATERIALS AND METHODS

Pertussis VCR were estimated among all women who gave birth and men who took paternity leave, in 2016 or 2017, from a 1/97th random sample of French claims data. Two distinct study periods were defined based on current recommendations for the cocooning strategy: the "common practice" and the "parental project" periods.

RESULTS

In 2016, the pertussis VCR of women having given birth and men having taken paternity leave was 47.2 and 47.1%, respectively (46.1 and 45.6% in 2017, respectively). About one quarter of vaccinations were performed during the "parental project" period, with the vaccine most frequently reimbursed during the month of childbirth for women (57.1% in 2016 and 49.4% in 2017) and before or during the month the paternity leave began for men (about 78% in both 2016 and 2017). General practitioners were the main prescribers in private practice, even during the "parental project" period.

CONCLUSION

To optimize the protection for infants, the main objective of the cocooning strategy, pertussis immunization coverage of adults and seniors needs to be improved. Moreover, cocooning vaccination linked to a parental project needs to be performed earlier, during pregnancy (for those around the mother) or in immediate post-partum (e.g., during the maternity stay).

Effectiveness of BNT162b2, mRNA-1273, and ChAdOx1-S vaccines against severe covid-19 outcomes in a nationwide mass vaccination setting: cohort study

Objective

To estimate the effectiveness of the three covid-19 vaccines by Pfizer-BioNTech (BNT162b2), Moderna (mRNA-1273), and Oxford-AstraZeneca (ChAdOx1-S) in people after receiving two doses.

Design

Cohort study.

Setting

Nationwide, population based data in France, from the French National Health Data System (Système National des Données de Santé), between 27 December 2020 and 30 April 2021.

Participants

Adults aged ≥50 years receiving a first dose of BNT162b2, mRNA-1273, or ChAdOx1-S were randomly selected (1:1) and matched on the date of vaccination with one unvaccinated control. Individuals were matched on year of birth, sex, region of residence, and residence in a nursing home (for individuals aged ≥75 years). All individuals were followed up until 20 August 2021.

Main outcome measures

Primary outcome measure was vaccine effectiveness estimated at least 14 days after the second dose against covid-19 related hospital admission using Cox proportional hazards models adjusted for baseline characteristics and comorbidities. Vaccine effectiveness against covid-19 related death in hospital was also investigated.

Results

11 256 832 vaccinated individuals were included in the study (63.6% (n=7 161 658) with the BNT162b2 vaccine, 7.6% (n=856 599) with the mRNA-1273 vaccine, and 28.8% (n=3 238 575) with the ChAdOx1-S vaccine), along with 11 256 832 matched unvaccinated controls. During follow-up (up to 20 August 2021), 43 158 covid-19 related hospital admissions and 7957 covid-19 related deaths in hospital were registered. Compared with unvaccinated controls, vaccine effectiveness of two doses against covid-19 related hospital admission was 91% (95% confidence interval 91% to 92%), 95% (93% to 96%), and 91% (89% to 94%) for the BNT162b2, mRNA-1273, and ChAdOx1-S vaccines, respectively. Similar results were observed for vaccine effectiveness of two doses against covid-19 related deaths in hospital (BNT162b2, 91% (90% to 93%); mRNA-1273, 96% (92% to 98%); and ChAdOx1 nCoV-19, 88% (68% to 95%)). At 5-6 months after receiving the second dose of vaccine, effectiveness remained high at 94% (92% to 95%) for the BNT162b2 vaccine and 98% (93% to 100%) for the mRNA-1273 vaccine. Vaccine effectiveness of ChAdOx1-S estimated at 3-4 months was 90% (63% to 97%). All three vaccines remained effective at the time of circulation of the delta variant of SARS-CoV-2 between 1 July and 20 August 2021 (effectiveness between 89% and 95%).

Conclusions

These findings provide evidence indicating that two doses of ChAdOx1-S is as effective as two doses of mRNA vaccines in France against the alpha and delta variants of SARS-CoV-2. The effectiveness of ChAdOx1-S should be further examined with a longer follow-up and in the light of the circulation of new SARS-CoV-2 variants of concern.

The dynamics of the general practitioner-nephrologist collaboration for the management of patients with chronic kidney disease before and after dialysis initiation: a mixed-methods study

Background:

Effective collaboration between general practitioners (GP) and nephrologists is crucial in CKD care. We aimed to analyse GPs’ and nephrologists’ presence and involvement in CKD care and assess how they intertwine to shape patients’ trajectories.

Methods:

We conducted a mixed-methods study that included all patients with CKD who started dialysis in France in 2015 (the REIN registry) and a sample of nephrologists and GPs. We quantified professionals’ presence through patients’ reimbursed healthcare from the French National Health Data System, 2 years before and 1 year after dialysis start. Involvement in CKD care was derived from the nephrologists’ and GPs’ interviews.

Results:

Among 8856 patients included, nephrologists’ presence progressively increased from 29% to 67% of patients with a contact during the 2 years before dialysis start. However, this was partly dependent on the GPs’ referral practices. Interviews revealed that GPs initially controlled the therapeutic strategy on their own. Although unease grew with CKD’s management complexity, reducing their involvement in favour of nephrologists, GPs’ presence remained frequent throughout the pre-dialysis period. Upon dialysis start, nephrologists’ presence and involvement became total, while GPs’ greatly decreased (48% of patients with a contact at month 12 after dialysis start). Collaboration was smooth when GPs maintained contact with patients and could contribute to their care through aspects of their specialty they valued.

Conclusions:

This mixed-methods study shows presences and forms of involvement of GPs and nephrologists in CKD care adjusting along the course of CKD and unveils the mechanisms at play in their collaboration.

Care management and 90-day post discharge mortality in patients hospitalized for myocardial infarction and COVID-19: A French nationwide observational study

BACKGROUND

Concomitant or cured coronavirus disease 2019 (COVID-19) in patients with myocardial infarction (MI) may lead to difficulties in acute care management and impair prognosis.

AIMS

To describe and compare the characteristics, care management and 90-day post discharge outcomes of patients hospitalized for MI who did not have COVID-19 with those of patients with concomitant or previous hospital-diagnosed COVID-19.

METHODS

This population-based French study included all patients hospitalized for MI in France (30 December 2019 to 04 October 2020) from the French National Health Data System. Outcomes were described for each COVID-19 group and compared using adjusted logistic regression analysis.

RESULTS

Among 55,524 patients hospitalized for MI, 135 had previous hospital-diagnosed COVID-19 and 329 had concomitant COVID-19. Patients with previous hospital-diagnosed COVID-19 had more personal history of cardiovascular diseases than those without concomitant/previous confirmed COVID-19. In-hospital and 90-day post discharge mortality rates of patients with previous COVID-19 were 8.1% and 4.0%, respectively, compared with 3.5% and 3.0% in patients without concomitant/previous confirmed COVID-19 (odds ratio [OR]_{adjin-hospital} 1.83, 95% confidence interval [CI] 0.97-3.46; OR_{adjpostdischarge} 0.77, 95% CI 0.28-2.13). Patients with concomitant COVID-19 had more personal history of cardiovascular diseases, but also a poorer prognosis than their no concomitant/no previous confirmed COVID-19 counterparts; they presented excess cardiac complications during hospitalization (OR_adj 1.62, 95% CI 1.29-2.04), in-hospital mortality (OR_adj 3.31, 95% CI 2.32-4.72) and 90-day post discharge mortality (OR_adj 2.09, 95% CI 1.24-3.51).

CONCLUSIONS

In-hospital and 90-day post discharge mortality of patients hospitalized for MI who had previous hospital-diagnosed COVID-19 did not seem to differ from those hospitalized for MI alone. Conversely, concomitant COVID-19 and MI carried a poorer prognosis extending beyond the hospital stay. Special attention should be given to patients with simultaneous COVID-19 and MI, in terms of acute care and secondary prevention.

Outcomes associated with antidepressant treatment according to the number of prescriptions and treatment changes: 5-year follow-up of a nation-wide cohort study

BACKGROUND

Naturalistic studies regarding clinical outcomes associated with antidepressant treatment duration have yielded conflicting results, possibly because they did not consider the occurrence of treatment changes. This nation-wide population-based study examined the association between the number of filled prescriptions and treatment changes and long-term psychiatric outcomes after antidepressant treatment initiation.

METHODS

Based on the French national health insurance database, 842,175 adults who initiated an antidepressant treatment in 2011 were included. Cox proportional-hazard multi-adjusted regression models examined the association between the number of filled prescriptions and the occurrence of treatment changes 12 months after initiation and four outcomes during a 5-year follow-up: psychiatric hospitalizations, suicide attempts, sick leaves for a psychiatric diagnosis, new episodes of antidepressant treatment.

RESULTS

During a mean follow-up of 4.5 years, the incidence rates of the four above-mentioned outcomes were 13.49, 2.47, 4.57, and 92.76 per 1,000 person-years, respectively. The number of filled prescriptions was associated with each outcome (adjusted HRs [95% CI] for one additional prescription ranging from 1.01 [1.00-1.02] to 1.10 [1.09-1.11]), as was the occurrence of at least one treatment change vs. none (adjusted HRs [95% CI] ranging from 1.18 [1.16-1.21] to 1.57 [1.79-1.65]). Furthermore, the adjusted HRs [95% CI] of the number of filled prescriptions were greater in patients with (vs. without) a treatment change for psychiatric hospitalizations (1.12 [1.11-1.14] vs. 1.09 [1.08-1.10], p for interaction = 0.002) and suicide attempts (1.12 [1.09-1.15] vs. 1.06 [1.04-1.08], p for interaction = 0.006).

LIMITATIONS

Lack of clinical data about the disorders warranting the prescriptions or their severity.

CONCLUSION

Considering treatment changes is critical when using administrative claims database to examine the long-term psychiatric outcomes of antidepressant treatments in real-life settings.

Sleep Apnea-Specific Hypoxic Burden, Symptom Subtypes, and Risk of Cardiovascular Events and All-Cause Mortality

Rationale: Data from population-based cohorts suggest that symptom subtypes and obstructive sleep apnea (OSA)-specific hypoxic burden (HB) could help to better identify patients with OSA at high cardiovascular (CV) risk. Objectives: We aimed to evaluate whether those new markers are associated with the risk of major adverse CV events (MACE) in clinical setting. Methods: Data from the Pays de la Loire cohort were linked to health administrative data to identify the occurrence of MACE (a composite outcome including all-cause mortality, acute myocardial infarction, stroke, and unplanned coronary revascularization) in patients with newly diagnosed OSA and no overt CV disease. Latent class analysis was used to identify subtypes based on eight clinically relevant variables. HB was defined as the total area under the respiratory event-related desaturation curve. Cox proportional hazards models were used to evaluate the association of symptom subtypes and HB with MACE. Measurements and Main Results: Four symptom subtypes were identified (minimally symptomatic [22.0%], disturbed sleep [17.5%], excessively sleepy [49.8%], and moderately sleepy [10.6%]). After a median follow-up of 78 months (interquartile range, 52-109), 592 (11.05%) of 5,358 patients experienced MACE. In a fully adjusted model, HB and overall nocturnal hypoxemia assessed by sleep time with oxygen saturation <90% were the only predictors of MACE (hazard ratio, 1.21; 95% confidence interval, 1.07-1.38; and hazard ratio, 1.34; 95% confidence interval, 1.16-1.55, respectively). The association appeared stronger toward younger patients and women. Conclusion: In clinical setting, patients with OSA who demonstrate elevated OSA-specific HB are at higher risk of a CV event and all-cause mortality. Symptom subtypes were not associated with MACE after adjustment for confounders.

Antipsychotic prescribing practices in real-life (APPREAL study): Findings from the French National Healthcare System Database (2007-2017)

BACKGROUND

Antipsychotics are used in a large variety of psychiatric and neurological disorders; investigating their use in real life is important to understand national prescribing practices, as well as to determine the levels of patient adherence.

METHODS

Using a 1/97e random sample (General Sample of Beneficiaries, EGB) of the French health insurance reimbursement database, we conducted a historical cohort study on the 2007-2017 period. The aim was to describe the sociodemographic characteristics of patients, the types of antipsychotics dispensed, the types of prescribers, the mean doses and average durations of treatment, the co-dispensed medications, and the levels of adherence to treatment. To exclude punctual uses of antipsychotics, we selected only patients with a continuous dispensing of the same antipsychotic over at least 3 months.

RESULTS

In total, 13,799 subjects (1.66% of the EGB sample) were included (56.0% females; mean age 55.8 ± 19.4 years). Risperidone (19.3%), cyamemazine (18.7%), olanzapine (11.9%), tiapride (8.8%), and haloperidol (7.5%) were the five most prescribed antipsychotics. 44.9% of prescriptions were written by general practitioners, 34.1% by hospital practitioners, and 18.4% by private-practice psychiatrists. On average, the mean dispensed doses were relatively low, but the variation range was large. Long-acting forms were used in 5.4% of the sample, and clozapine in 1.3%. 34.2% of patients received more than one antipsychotic, and almost 15% were prescribed at least three concomitant antipsychotics. Paliperidone and clozapine were associated with the highest levels of adherence, and risperidone and haloperidol with the lowest ones.

CONCLUSION

An important heterogeneity of antipsychotic prescribing practices was observed in France. The rate of use of long-acting antipsychotics was low, whereas multiple antipsychotic prescriptions were frequent.

Impact of extending direct antiviral agents (DAA) availability in France: an observational cohort study (2015-2019) of data from French administrative healthcare databases (SNDS)

Background

Direct antiviral agents (DAAs) became available in France in 2014 for the treatment of chronic hepatitis C (CHC) in patients with severe fibrosis (prioritized access); in 2017, DAAs became available to all CHC patients (universal access). We evaluated the impact of extending DAA availability on CHC patient care, especially on screening and time to treatment.

Methods

Adult patients affiliated with the national health insurance system (SNDS) who were screened or treated for CHC between 2015 and 2019 were included. Algorithms were developed to identify at-risk subpopulations.

Findings

The proportion of screened patients increased by 1% between 2015 and 2019, from 4·6% to 5·6%. The main nonexclusive risk factors for CHC were psychiatric conditions (27%), drug use (21%) and HIV positivity (11%); more than 50% of psychiatric patients had additional risk factors, mainly drug use with a 38% to 52% overlap.

The median interval between the last screening test and treatment initiation decreased from 64 days in 2015 to 37 days in 2019.

During the study period, 71,466 patients began CHC treatment (median age 55 [48-62]; 59% male), including 46% of “at-risk” patients with an increase in treatment initiation by 44% between 2015 and 2017 and a decrease of 46% between 2017 and 2019. Only 2,212 (3%) patients were treated at least twice.

Among treated patients, the proportion of HIV+ patients decreased from 19% to 8% (prioritization consequence), while the proportions increased in the other at-risk subpopulations.

Interpretation

we showed that policies extending DAA availability are associated with a screening increase and a decrease in the time to treatment initiation, while universal access led to a surge in treatment initiations in 2017. This study may also contribute to improving the cascade of care in the at-risk subpopulations. For instance, by pointing out their relative importance, especially for the psychiatric subpopulation, it highlights the importance to address them with tailored policies.

Incidence and risk of cancer among Multiple Sclerosis patients: a matched population-based cohort study

BACKGROUND

Studies have not yet found conclusive results on the risk of cancer in patients with Multiple Sclerosis (MS). This study aimed to compare the incidence of all cancers and of specific types of cancer between MS patients and the general population by age and by sex.

METHODS

All prevalent MS patients identified between 2008-2014 in the nationwide French healthcare database (SNDS) and without history of malignancy, were included in a cohort study followed-up until cancer occurrence, date of death or 31 December 2015, whichever came first. MS patients were matched based on sex and year of birth, to non-MS controls from the general population without cancer before index date. Incidence rate was reported per 100,000 person-year (PY) and risk of cancer was estimated by type of cancer, age and sex using a Cox model (Hazard Ratio, HRs and its 95% confidence intervals, 95%CI).

RESULTS

Overall, 576 cancers per 100,000 PY were observed in MS patients versus 424 per 100,000 PY in the control population. The risk of cancer was higher among MS patients than among population controls whether considered overall (HR: 1.36, 95%CI: 1.29-1.43) or for prostate (HR: 2.08, 95%CI: 1.68-2.58), colorectal and anal (HR: 1.35, 95%CI: 1.16-1.58), trachea bronchus and lung (HR: 2.36, 95%CI: 1.96-2.84), and to a lesser extent, breast cancer (HR: 1.12, 95%CI: 1.03-1.23).

CONCLUSION

MS patients were associated with increased risk of cancer compared to population controls.

Leveraging National Claims and Hospital Big Data: Cohort Study on a Statin-Drug Interaction Use Case

Background

Linking different sources of medical data is a promising approach to analyze care trajectories. The aim of the INSHARE (Integrating and Sharing Health Big Data for Research) project was to provide the blueprint for a technological platform that facilitates integration, sharing, and reuse of data from 2 sources: the clinical data warehouse (CDW) of the Rennes academic hospital, called eHOP (entrepôt Hôpital), and a data set extracted from the French national claim data warehouse (Système National des Données de Santé [SNDS]).

Objective

This study aims to demonstrate how the INSHARE platform can support big data analytic tasks in the health field using a pharmacovigilance use case based on statin consumption and statin-drug interactions.

Methods

A Spark distributed cluster-computing framework was used for the record linkage procedure and all analyses. A semideterministic record linkage method based on the common variables between the chosen data sources was developed to identify all patients discharged after at least one hospital stay at the Rennes academic hospital between 2015 and 2017. The use-case study focused on a cohort of patients treated with statins prescribed by their general practitioner or during their hospital stay.

Results

The whole process (record linkage procedure and use-case analyses) required 88 minutes. Of the 161,532 and 164,316 patients from the SNDS and eHOP CDW data sets, respectively, 159,495 patients were successfully linked (98.74% and 97.07% of patients from SNDS and eHOP CDW, respectively). Of the 16,806 patients with at least one statin delivery, 8293 patients started the consumption before and continued during the hospital stay, 6382 patients stopped statin consumption at hospital admission, and 2131 patients initiated statins in hospital. Statin-drug interactions occurred more frequently during hospitalization than in the community (3800/10,424, 36.45% and 3253/14,675, 22.17%, respectively; P<.001). Only 121 patients had the most severe level of statin-drug interaction. Hospital stay burden (length of stay and in-hospital mortality) was more severe in patients with statin-drug interactions during hospitalization.

Conclusions

This study demonstrates the added value of combining and reusing clinical and claim data to provide large-scale measures of drug-drug interaction prevalence and care pathways outside hospitals. It builds a path to move the current health care system toward a Learning Health System using knowledge generated from research on real-world health data.

Five-year cause-specific survival after meningioma surgery. A nationwide population-based study

BACKGROUND

Survival after meningioma surgery is often reported with inadequate allowance for competing causes of death.

METHODS

We processed the French administrative medical database (Système National des Données de Santé: SNDS), to retrieve appropriate cases of surgically treated meningioma. Cause-specific survival in meningioma-related death was analyzed with the Fine & Gray (F&G) and cause-specific (CS) Cox models to identify associated factors.

RESULTS

Five-year cumulative incidence was 2.85% for meningioma-related death and 6.3% for unrelated death (P<0.001). In the adjusted F&G and cause-specific Cox regression models for meningioma-related death, gender, age at surgery, co-morbidities, neurofibromatosis type 2, tumor insertion, tumor grade, cerebrospinal fluid (CSF) shunt insertion, preoperative embolization and need for redo surgery for recurrence emerged as independent prognostic factors of cause-specific survival (CSS) in meningioma-related death.

CONCLUSION

At 5 years, the risk of meningioma-unrelated death was 2.21-fold greater than the risk of dying from the meningioma disease. Five-year CSS after meningioma surgery was greater in younger adults with benign spinal meningioma with low comorbidity. Those with malignant cranial tumor requiring preoperative embolization or CSF shunting for associated hydrocephalus and with severely degraded overall health status showed a significantly increased risk of meningioma-related death. Redo surgery for recurrence failed to improve the risk of meningioma-related death. We recommend the use of net survival methods such as CSS in meningioma studies where unrelated mortality is predominant, as this approach results in more accurate estimates of disease risk and associated predictors.

Impact of recommendations on sodium valproate prescription among women with epilepsy: An interrupted time-series study

BACKGROUND

The European Medicines Agency (EMA) has developed risk minimization measures (RMMs) to reduce the use of the teratogenic drug, sodium valproate (VPA). The objective was to assess the impact of these RMMs among females with epilepsy in France.

METHODS

We used data from the French National Health Insurance Database (SNDS), including 114,936 females aged under 50, with a reimbursement claim for an antiepileptic drug from January 2011 to December 2017, and identified as people with epilepsy. We used a controlled interrupted time series stratifying on age: girls (0-14 years old) and women of childbearing age (15-49 years), and with 129,917 males as controls.

RESULTS

VPA prevalent use among girls and women of childbearing age with epilepsy decreased significantly after the issue of the RMMs (trend changes of, respectively, -5 and -4 users per 1000 females at-risk per quarter in comparison to the control group). We did not detect any significant change in VPA incident use.

CONCLUSIONS

VPA use decreased over the study period among females with epilepsy but there were still 317 women and 206 girls started on VPA therapy VPA in 2017 (8 per 1000 at-risk and 18 per 1000, respectively). This suggests that either the measures should be strengthened or that the lowest level of VPA use has been reached. In this context, the introduction of a new RMM (in 2018) needs to be evaluated.

Myocardial infarction throughout 1 year of the COVID-19 pandemic: French nationwide study of hospitalization rates, prognosis and 90-day mortality rates

Background

Studies reported a decrease in hospital admissions for myocardial infarction (MI) in early 2020 as a result of the coronavirus disease 2019 (COVID-19) crisis, mainly restricted to the beginning of the pandemic.

Aims

To describe national trends in hospital admissions for MI in 2020, and to compare patient characteristics, in-hospital prognosis and 90-day mortality between patients who had an MI in 2020 and those admitted in 2017–2019.

Methods

All patients hospitalized for MI in France from 2017 to 2020 were selected from the national hospital discharge database. Analyses compared temporal trends in MI admissions, in-hospital cardiac complications and mortality rates in 2020 versus 2017–2019.

Results

In 2020, 94,747 patients were hospitalized for MI, corresponding to a 6% decrease in MI admissions compared with 2017–19. This decrease was larger during the first lockdown (–24%; P < 0.0001) than during the second lockdown (–8%; P < 0.0001). Reductions in MI admissions were more pronounced and longer among patients with non-ST-segment elevation MI, older people and women. An increase in ST-segment elevation MI admissions was observed between lockdowns (+4%; P = 0.0005). Globally, and after adjustment for age, sex and calendar year, in-hospital and 90-day post-discharge mortality rates did not differ in 2020 versus 2017–19: incidence rate ratio (IRR)_{adjin-hospital} 1.03, 95% confidence interval (CI) (0.98–1.08); IRR_{adj90-daypost-discharge} 1.06, 95% CI (0.98–1.13).

Conclusions

In 2020, a significant decrease in MI admissions was observed, and was marked at the beginning of the year. This highlights the need to disseminate public information on the importance of maintaining care and regular medical follow-up. The effect of the COVID-19 crisis on acute and 3-month outcomes of patients hospitalized for MI appears limited. Nevertheless, monitoring of chronic MI complications and the impact on non-hospitalized patients should continue.

Healthcare costs in implantable cardioverter-defibrillator recipients: A real-life cohort study on 19,408 patients from the French national healthcare database

BACKGROUND

The aim is to report healthcare costs in a nationwide cohort of patients with an implantable cardioverter defibrillator (ICD).

METHODS

This real-life longitudinal retrospective cohort study was based on the French National Health Data System and enrolled all adult patients from the general health insurance scheme implanted with an ICD between 2008 and 2011, and followed them until 2018.

RESULTS

Overall, 19,408 patients were included (mean age 63.8, SD 12.4 years, 81.6% males), with cardiac resynchronization therapy (CRTD), single-chamber, and dual-chamber ICD in 42.5%, 29.8%, 27.7% of patients, respectively. After a mean follow-up of 6.6 SD 3.3 years, 9514 patients (49.0%) died, and 8678 patients (44.7%) had their ICD replaced. The total healthcare cost (all diseases and injuries combined) was €15,893/patient-year, of which 32% were estimated to be ICD-related. These ICD-related costs were: the implantation hospital stay (representing 59% of the ICD-related costs), ICD replacement (22%), complications' management (11%), and follow-up (9%). Some health events (e.g., a complication during ICD replacement) were counted in two categories, hence the sum of the proportions is >100%. Being under 55 vs. above 75 years old, being treated for hypertension vs. not treated, and receiving a CRT-D vs. a single-chamber ICD each increased the mean total ICD-related cost per patient by approximately 20%; ICD replacement vs. no replacement increased it by 71%.

CONCLUSIONS

Almost two thirds of the total ICD patients' healthcare costs remained not ICD-related. Advancing the understanding of direct and indirect costs may help improving cost-effectiveness of patients' care pathway.

Risk of bleeding associated with nonsteroidal anti-inflammatory drug use in patients exposed to antithrombotic therapy: a case-crossover study

Concomitant nonsteroidal anti-inflammatory drug (NSAIDs) and antithrombotic drug use is associated with an increased risk of bleeding, mainly gastrointestinal. The goal of this study was to quantify the transient increase in the risk of hospitalization for bleeding associated with NSAID use in patients treated with antiplatelet agents or anticoagulants. We performed an unidirectional case-crossover study using the EGB (Échantillon généraliste de bénéficiaires), a permanent random sample of the French nationwide health database. Patients receiving antithrombotic therapy and hospitalized for bleeding between 2009 and 2017 were included. We compared their NSAID exposure during a 15-day hazard window immediately prior to hospital admission to three earlier 15-day control windows. The risk of hospitalization for bleeding associated with the recent use of NSAIDs was estimated using conditional logistic regression to estimate odds ratios. During the study period, 33 patients treated with anticoagulants and 253 treated with antiplatelet agents received NSAIDs and were included in the case-crossover analysis. We found an increased risk of hospitalization for gastrointestinal bleeding after exposure to NSAIDs with an adjusted OR of 3.59 (95%CI, 1.58;8.17) in patients receiving anticoagulant therapy and 1.44 (95%CI, 1.07;1.94) in patients receiving antiplatelet therapy. The risk of non-gastrointestinal bleeding was also increased after exposure to NSAIDs with an adjusted OR of 2.72 (95%CI, 1.23;6.04) in patients exposed to anticoagulant therapy. The risk of gastrointestinal and non-gastrointestinal bleeding increases after NSAID use in patients treated with anticoagulants, while the risk of gastrointestinal bleeding increases, but to a lesser extent in those treated with antiplatelets. This article is protected by copyright. All rights reserved.

Outcome After Protontherapy for Progression or Recurrence of Surgically Treated Meningioma

Background

To assess the outcome after meningioma surgery and protontherapy (PT).

Methods

We processed the French Système National des Données de Santé database to retrieve appropriate cases of meningiomas operated and irradiated between 2008 and 2017. Survival methods were implemented.

Results

One hundred ninety-three patients who received PT after meningioma surgery over a 10-year period were identified. Of the 193 patients, 75.6% were female. Median age at surgery was 50 years (interquartile range [IQR] 41–62). The median number of PT fractions was 31 (IQR 30–39) given over a median duration of 52 days (IQR 44–69). Fourteen patients (7.3%) also received photon radiotherapy and six patients (3.1%) stereotactic radiosurgery. Median follow-up was 4.4 years (IQR 3.86–4.71). Five-year progression-free survival (PFS) rate was 69% (95% confidence interval [CI] 62.1–76.6). For benign, atypical, and malignant meningioma, 5-year PFS rates were 71.5% (95% CI 64.4–79.4), 55.6% (95% CI 32.5–95), and 35.6% (95% CI 12.8–98.9), respectively (p<0.01). In the adjusted regression, tumour location (hazard ratio [HR]=0.1, 95% CI 0.05–0.22, p<0.001), aggressive meningioma (HR=2.26, 95% CI 1.1–4.66, p=0.027), and the need of cerebrospinal fluid (CSF) insertion for hydrocephalus (HR=3.51, 95% CI 1.32–9.31, p=0.012) remained significantly associated to the PFS. All grades considered, 5-year overall survival (OS) rates was 89.7% (95% CI 84.6–95.1). For benign, atypical, and malignant meningioma, 5-year OS rates were 93% (95% CI 88.7–97.4), 76.4% (95% CI 51.4–100), and 44.4% (95% CI 16.7–100), respectively (p<0.01). In the multivariable regression, an older age above 70 years (HR=5.95, 95% CI 2.09–16.89, p<0.001) associated to a high level of comorbidities (HR=5.31, 95% CI 1.43–19.78, p=0.013) and a malignant meningioma (HR=5.68, 95% CI 1.54–20.94, p=0.009) remained significantly associated to a reduced OS.

Conclusion

Five-year PFS and OS after meningioma surgery and PT is favourable but impaired for older patients with high level of morbidities, tumour of the convexity, malignant histopathology and for those requiring CSF shunting. Further inclusion and prolonged follow-up is required to assess other predictors such as sex, tumour volume, or given dose.

Changes in the epidemiology of patients hospitalized in France with deep venous thrombosis and pulmonary embolism during the COVID-19 pandemic

INTRODUCTION

The onset of the COVID-19 pandemic and the first national lockdown implemented might have disrupted the epidemiology of deep venous thrombosis (DVT) and pulmonary embolism (PE). This study aimed to analyze time trends in patients hospitalized for DVT and PE in France and related in-patient and 90-day post-admission mortality rates.

MATERIALS AND METHODS

All patients hospitalized in France for DVT or PE between January and September (weeks 1-40) for each year from 2017 to 2020, were selected. Weekly incidence rate ratios (IRR) were computed to compare the rates of patients hospitalized in 2020 with those hospitalized in 2017-2019.

RESULTS

Compared with the 2017-2019 study period, the rates of patients hospitalized with a primary diagnosis (PD) of DVT or PE in 2020 were significantly (50 and 40%, respectively) lower during weeks 12-13. The rate of patients hospitalized with an associated diagnosis (AD) of PE during weeks 12-19 of 2020 was twice as high as in the same period in 2017-2019. The prevalence of COVID-19 in patients hospitalized with a PD of DVT and PE, and in those hospitalized with an AD of DVT and PE reached respectively 4.0, 9.6, 17.2 and 44.6 during the country's first lockdown. Inpatients case-fatality rates in patients hospitalized with an AD of PE increased significantly during weeks 12-13.

CONCLUSIONS

Epidemiology of VT and PE was seriously impacted by the COVID-19 pandemic in 2020 in France, with a significant decrease in the rate of patients hospitalized for PE and a threefold increase in the related in-patient mortality rate. This highlight the need to inform the general population about the symptoms of PE and about the need to immediately seek medical care, particularly those infected with SARS-CoV-2.

Cardiovascular risk and mortality prediction in patients suspected of sleep apnea: a model based on an artificial intelligence system

Objective. Cardiovascular disease (CVD) is one of the leading causes of death worldwide. There are many CVD risk estimators but very few take into account sleep features. Moreover, they are rarely tested on patients investigated for obstructive sleep apnea (OSA). However, numerous studies have demonstrated that OSA index or sleep features are associated with CVD and mortality. The aim of this study is to propose a new simple CVD and mortality risk estimator for use in routine sleep testing.Approach. Data from a large multicenter cohort of CVD-free patients investigated for OSA were linked to the French Health System to identify new-onset CVD. Clinical features were collected and sleep features were extracted from sleep recordings. A machine-learning model based on trees, AdaBoost, was applied to estimate the CVD and mortality risk score.Main results. After a median [inter-quartile range] follow-up of 6.0 [3.5-8.5] years, 685 of 5234 patients had received a diagnosis of CVD or had died. Following a selection of features, from the original 30 features, 9 were selected, including five clinical and four sleep oximetry features. The final model included age, gender, hypertension, diabetes, systolic blood pressure, oxygen saturation and pulse rate variability (PRV) features. An area under the receiver operating characteristic curve (AUC) of 0.78 was reached.Significance. AdaBoost, an interpretable machine-learning model, was applied to predict 6 year CVD and mortality in patients investigated for clinical suspicion of OSA. A mixed set of simple clinical features, nocturnal hypoxemia and PRV features derived from single channel pulse oximetry were used.

Variation in end-of-life care and hospital palliative care among hospitals and local authorities: A preliminary contribution of big data

BACKGROUND

Many studies explore the clinical and ethical dimensions of care at the end-of-life, but fewer use administrative data to examine individual and geographic differences, including the use of palliative care.

AIM

Provide a population-based perspective on end-of-life and hospital palliative care among local authorities and hospitals in France.

DESIGN

Retrospective cohort study of care received by 17,928 decedents 65 and over (last 6 months of life), using the French national health insurance database.

RESULTS

55.7% of decedents died in acute-care hospitals; 79% were hospitalized in them at least once; 11.7% were admitted at least once for hospital palliative care. Among 31 academic medical centers, intensive care unit admissions ranged from 12% to 67.4%; hospital palliative care admissions, from 2% to 30.6%. Across local authorities, for intensive care unit days and hospital palliative care admissions, the ratios between the values at the third and the first quartile were 2.4 and 1.5. The odds of admission for hospital palliative care or to an intensive care unit for more than 7 days were more than twice as high among people ⩽85 years (aOR = 2.11 (1.84-2.43) and aOR = 2.59 (2.12-3.17), respectively). The odds of admission for hospital palliative care were about 25% lower (p = 0.04) among decedents living in local authorities with the lowest levels of education than those with the highest levels.

CONCLUSION

The variation we document in end-of-life and hospital palliative care across different categories of hospitals and 95 local authorities raises important questions as to what constitutes appropriate hospital use and intensity at the end-of-life.

Estimation of the Width of Uncertainty in Care Consumption and Costs When Using Common Data Collection Tools in Economic Evaluations: A Benchmark for Sensitivity Analyses

OBJECTIVES

This study aimed to evaluate the uncertainty related to the use of common collection tools to assess costs in economic evaluations compared with an exhaustive administrative database.

METHODS

A pragmatic study was performed using preexisting cost-effectiveness studies. Patients were probabilistically matched with themselves in the French National Health Data System (Système National des Données de Santé [SNDS]), and all their reimbursed hospital and ambulatory care data during the study were extracted. Outcomes included the ratio of the number of each type of resources consumed using trial data (case report forms for ambulatory care and local hospital data for hospital care) versus the SNDS and the ratio of corresponding costs. Mean ratios and 95% confidence intervals (CIs) were calculated using bootstrapping. The impact of the collection tool on the result of the economic evaluation was calculated with the difference in costs between the 2 treatment arms with both collection methods.

RESULTS

Five cost-effectiveness studies were included in the analysis. A total of 397 patients had the SNDS hospital data, and 321 had ambulatory care data. Common collection tools underestimated hospital admissions by 13% (95% CI 8-20), corresponding costs by 5% (95% CI 2-14), and ambulatory acts by 41% (95% CI 33-51), with large variations in costs depending on the study. There was no change in the economic conclusion in any study.

CONCLUSIONS

The use of common collection tools underestimates healthcare resource consumption and its associated costs, particularly for ambulatory care. Our results could provide useful evidence-based estimates to inform sensitivity analyses' parameters in future cost-effectiveness analyses.