Since 2013, the National Health Care Institute in the Netherlands has systematically analyzed the appropriateness of care provided under public health insurance. Here we present the method used, the results up to now, and what we have learned from it.

The appropriate care program consists of four phases: screening, in-depth analysis, implementation, and evaluation. Stakeholder involvement is a central part of the process. For every ICD-10 area, a screening took place to select care trajectories for in-depth analysis with a potential for wiser choices and more appropriate care. The in-depth analysis indicates which improvements can be made to reach more appropriate care, by assessing guideline adherence. During the implementation phase, which is primarily carried out by clinicians, patients and health insurers, actions are taken to improve care on the identified points. In the evaluation phase, we examine to what extent improvements have been achieved.

Currently, all ICD-10 areas have been screened and 29 selected care trajectories have been subjected to in-depth analyses. The analyses resulted in the identification of more than a hundred areas for potential improvement of the appropriateness of care. For most topics implementation of changes is currently taking place. The four most important impact-enhancing lessons learned by applying the working method are: (i) ICD-10 areas as a starting point for screening are not the most efficient method to reach the biggest impact. (ii) The screening should take a societal perspective. (iii) All public and private parties involved should fulfill their role and take responsibility. (iv) To fulfill our own role better, the working method should be more connected to health technology assessment for reimbursement decisions.

The program has resulted in the identification of many valuable points for improvement which could lead to more appropriate care in the coming years. The impact of the program could be increased through priority setting from a societal perspective and improving the connection to our other health technology assessment processes.

As part of the cyclic Appropriate Care programme of the National Health Care Institute in the Netherlands, a systematic analysis of hearing health care is taking place. Parties in hearing health care are actively involved throughout the entire process. This abstract focuses on lessons learned from the cooperation as a HTA body with a diverse group of stakeholders.

We carried out an in-depth analysis for the patient journey of both children and adults with ear complaints or hearing impairment. Different kinds of information were included in the analyses, including claims data, quantitative and qualitative research, analyses of (international) guidelines and patient information. A range of strategies were used to co-operate and interact with patient organizations, hearing health care professionals, institutes/hospitals and insurance companies.

Close collaboration between the project team and patient organizations turned out to be effective to comprehend patient’ perspectives. Data analyses were often found to be challenging in hearing health care, as the reimbursement data lacked sufficient information. In several cases, building bridges between parties, but also in relation to our HTA body was needed. Conclusions from the analyses were being shared and discussed with a panel of involved stakeholders, leading to support, but not always consensus on potential room for improvement. An internal review process turned out to be helpful in sharing experiences on effective multi-stakeholder management.

We believe that the process did influence the way stakeholders think about the appropriate use of the different available treatment options. Building bridges, and combining different perspectives from patient organizations, health-care professionals and insurance companies is necessary in a cyclic approach. The cyclic appropriate Care programme proved to be a constructive approach for collaboration with stakeholders.

Studies on neighbourhood characteristics and depression show equivocal results.

This large-scale pooled analysis examines whether urbanisation, socioeconomic, physical and social neighbourhood characteristics are associated with the prevalence and severity of depression.

Cross-sectional design including data are from eight Dutch cohort studies (n= 32 487). Prevalence of depression, either DSM-IV diagnosis of depressive disorder or scoring for moderately severe depression on symptom scales, and continuous depression severity scores were analysed. Neighbourhood characteristics were linked using postal codes and included (a) urbanisation grade, (b) socioeconomic characteristics: socioeconomic status, home value, social security beneficiaries and non-Dutch ancestry, (c) physical characteristics: air pollution, traffic noise and availability of green space and water, and (d) social characteristics: social cohesion and safety. Multilevel regression analyses were adjusted for the individual's age, gender, educational level and income. Cohort-specific estimates were pooled using random-effects analysis.

The pooled analysis showed that higher urbanisation grade (odds ratio (OR) = 1.05, 95% CI 1.01–1.10), lower socioeconomic status (OR = 0.90, 95% CI 0.87–0.95), higher number of social security beneficiaries (OR = 1.12, 95% CI 1.06–1.19), higher percentage of non-Dutch residents (OR = 1.08, 95% CI 1.02–1.14), higher levels of air pollution (OR = 1.07, 95% CI 1.01–1.12), less green space (OR = 0.94, 95% CI 0.88–0.99) and less social safety (OR = 0.92, 95% CI 0.88–0.97) were associated with higher prevalence of depression. All four socioeconomic neighbourhood characteristics and social safety were also consistently associated with continuous depression severity scores.

This large-scale pooled analysis across eight Dutch cohort studies shows that urbanisation and various socioeconomic, physical and social neighbourhood characteristics are associated with depression, indicating that a wide range of environmental aspects may relate to poor mental health.

None.