Study design
To explore the social and health situation of homeless people in Berlin a retrospective analysis of medical records from the health centre for homeless people of the Jenny De la Torre Foundation will be carried out. An overview of the study design can be obtained from Figure 1.
Data collection
The study centre provides donation-based and low-threshold medical care for homeless people. The centre includes a medical practice with medical specialists from various areas, a dental practice and an ophthalmic practice. Further, the centre offers psychological and social counselling as well as legal advice within various fields. In addition, two meals per day, clothing and a hairdresser are at the disposal of patients. The centre also offers the service of taking passport photographs. Every individual visiting the health centre has to first check in with the doctor’s office. At the first visit a medical record is generated for each person. Every time individuals visit the health centre, the reason of the visit and utilisation of any service is documented in the medical record. If required, patients may be treated anonymously.
As part of a preliminary screening we screened random selected records for content (N = 50). We summarised contained information and systematised the content under consideration of previous research (49, 50). As a result, eight different data sources were identified which can be assessed from medical records (see Table 1): 1) floating medical record (a file that contains all information available), 2) social and medical anamnesis (social and medical history of the patient; a standardised form that is filled out by the medical personnel at the first visit), 3) patient information sheet (a standardised form that contains information on any change of the homelessness, health insurance status, as well as the receipt of financial support), 4) doctors letters (including reports and hospital discharge letters), 5) psychological report (patients who make use of the psychological counselling in the centre are documented and a report of the counselling is written by the psychologist), 6) psychiatric record (a medical record about the psychological well-being of a patient), 7) documents provided by the health centre (doctors certificates or sick certificates), and 8) others (any other data available from the medical record). Based on the structure of data sources, a database was created using the software Research Electronic Data Capture (REDCap). REDCap is a browser-based, metadata-driven software for designing research databases. It is secured under data protection law by the Charité - Universitätsmedizin Berlin. The database is accessible online which simplifies the entry of data on site while ensuring data safety and protection. No personal identifying data is collected. During data collection, the survey is pseudonymised. For analysis and presentation of results, anonymised data is used. Data extraction started in July 2020 and will end in July 2022.
Inclusion and exclusion criteria
Every medical record created in the health centre for homeless people from the year 2006 until 2020 will be entered in the database and included in the analysis (N ≈ 3 000). There are no other inclusion or exclusion criteria.
Measures
A variety of health and social information can be obtained from the medical records. Generally, data can be divided in eight data sources. An overview of all data sources can be obtained from Table 1.
Pretest and quality assurance
After creating a first database, a pretest was conducted to ensure a complete and fully operational data entry. The pretest included N = 55 medical records. Data was entered by two experienced research associates. Data entry was documented and conspicuous differences and needs for adjustment collected. Based on the pretest the database was revised and edited. The new database was again tested with N = 10 medical records. Further, minor changes where implemented, and the data base was finalised. Data will be entered by two trained medical students. For quality insurance, prior to the beginning of data entry, a training on the data base and data collection was conducted with the medical students. Further, a code book for data entry was developed. Regular supervisions and random controls of data entry will be taking place during the entire period of data entry.
Statistical analysis
Total and cross-sectional sample characteristics will be presented in a descriptive analysis using Chi-Square Test, Mann-Whitney-U-Test or independent t-Test as appropriate to test (sub)group differences. The absolute standardised mean difference (ASMD) will be calculated to check the balancing of the characteristics. We consider ASMD<0.1 as adequate balance between groups. The level of significance will be considered at 0.05. Further, statistical analyses might use finite mixture modelling – particularly latent class analysis – in order to distinguish different types of social and health conditions. We will conduct latent variable regression in order to identify sociodemographic and disease-related factors associated with decreasing health conditions.
Program governances and ethical approval
The research project will be conducted according to the principles of Good Clinical Practice and the Declaration of Helsinki and was prospectively registered with the German Clinical Trials Register (trial registration number: DRKS00021172). The study received Human Research Ethics Committee approval from Charité - Universitätsmedizin Berlin (EA1/058/20). Further, the GIG study established an independently chaired steering committee prior to the start of data entry. It is comprised of representatives, program funders, board members, service providers engaged in the program, state government and the lead Chief Investigator of the research team. The committee members receive on a regular base short briefings from the research team on the current state of the project and in return, give advice on the research project. Additionally, the committee meets once a year in order to present and discuss the current state of the project. The first meeting already took place via a conference call in June 2020. The next chaired steering meeting will take place in spring 2021.