Background

Germany is one of the richest countries in the world and has comprehensive social security and medical care systems. At the same time, there are considerable social inequalities, which are also reflected in the health and life expectancy of the population. In Hamburg, too, there is a clear correlation between social class and health status. Socially disadvantaged population groups are often concentrated in segregated urban districts. The districts of Billstedt and Horn are among the poorest districts in Hamburg and are characterized by an above-average number of socio-economically disadvantaged residents. These problems are often exacerbated by a shortage of outpatient care services close to home. This can lead to inefficiencies in care and rising service costs. The problem is further exacerbated by the fact that services in the German healthcare system are not geared towards the needs of socially disadvantaged population groups.

The aim of the INVEST Billstedt/Horn project was to improve the problem situation outlined above by developing and implementing a range of care services that ensures low-threshold care close to home and without language barriers, thereby achieving the goals formulated in § 70 SGB V - quality, humanity and cost-effectiveness - in the socially disadvantaged Billstedt/Horn metropolitan region. To this end, a regional, integrated and population-oriented care model was developed and implemented, which focuses on prevention, health promotion and health maintenance. A unique focus of the project in Germany was the establishment of a health kiosk, a population-oriented, low-threshold and supportive district institution in which insured persons receive holistic advice, support and training on all health issues in their native language.

The project was scientifically monitored and evaluated by the Hamburg Center for Health Economics (HCHE). Based on quantitative and qualitative analyses, the achievement of objectives and the transferability of this innovative care approach were assessed and the success factors for successful implementation were analyzed. You can find a publication of the short evaluation report and the evaluation report in accordance with No. 14.1 ANBest-IF here.
The project "Hamburg Billstedt/Horn as a prototype for integrated full health care in deprived metropolitan regions"
(INVEST BILLSTEDT/HORN) was funded by the Joint Federal Committee (G-BA) with resources from the Innovation Fund (funding reference: 01NVF16025).

Objective

The follow-up project INVEST Billstedt/Horn II pursues two central objectives:

1. Causal evidence on medium-term effects: Investigation of the effects of the new form of care (NVF) over five years on service utilization and service expenditure.

2. Analysis of the heterogeneity of treatment effects: Identification of how effects vary according to patient characteristics, intensity of use and time course.

The analysis of heterogeneity shows which patients benefit most from the NVF and provides important information for more individualized, effective interventions. The results support decision-makers in further developing the NVF in a targeted manner, scaling it efficiently and transferring it sustainably to other regions.

The measurement of heterogeneous treatment effects (HTEs) of new integrated forms of care is necessary in view of scarce resources and funding gaps. The identification of subgroups with different effects enables more efficient targeting of interventions and strengthens targeted care on the basis of evidence-based decisions in care practice.

Research design/procedure

A quasi-experimental pretest-posttest design with a control group is used to investigate the effects of the new form of care (NPF). Differences before and after the introduction of the NVF are compared between intervention and control groups in order to determine the treatment effect. To minimize distortions due to confounding factors, the groups are made comparable by means of propensity score matching using a superlearner approach.

The temporal dynamics of the effects are analyzed using an event-study DiD approach. This approach tests the assumption of parallel trends before the intervention and records how the effects of the NVF develop over time.

The heterogeneity of the treatment effects is examined both according to patient characteristics and intensity of use. Subgroup analyses and causal machine learning methods such as double machine learning and causal forests are used for this purpose. The algorithmic approach makes it possible to identify previously unrecognized subpopulations for which the NVF is particularly effective - without having to specify them beforehand.