Marcel Slockers

54 Chapter 3 It should also be pointed out that under-registration might occur more often in a street doctor’s patient population than in a regular GP practice. In particular, regarding psy- chological disorders and addiction issues, diagnosing is rather complex. 4,22 Although under-registration may occur in a regular GP practice, the continuity of healthcare in this setting improves the quality of the record-keeping. Street doctors agreed upon how to register according to the ICPC, but a structural procedure of mutual alignment is still absent. There were no recorded changes in blood pressure checks or treatment. Another cause of under-reporting may be the fact that 29.0% of the homeless people have an intel- lectual disability. 23,24,25 We used the ten main categories that we used previously in our mortality study. By sampling, we checked whether large numbers were being overlooked with this group of ten categories. Only 8.0% had no ICPC registration across these ten main categories. It was not possible to conduct a comparative analysis with data from the NIVEL (a Dutch independent research institute for Dutch and European healthcare), as the number of person years of homeless people within a certain time span was unknown. By looking at the proportion of patients who visited the consultation hours at least once with a particular health problem, we were able to compare how homeless people in Rotterdam avail of primary healthcare as compared to non-homeless people in the city. However, we were only able to use the data of one regular GP practice. Vektis regional aggregate data indicate that this particular practice does not deviate or hardly deviates from the average GP practice in Rotterdam. (Vektis is a business intelligence centre specialising in healthcare). 26 The percentage of low-income patients of this practice is 40.0% as is the case in the Rotterdam region. The age profile of patients is also comparable, and there are only minor differences in total healthcare costs per registered patient. Similar to the mortality study in Rotterdam ten years ago, 7,8 we decided to compare homeless people to a generic group of non-homeless people in Rotterdam. There is a large group of vulnerable people with many problems at risk of homelessness. Therefore, further research into differences between healthcare for homeless and disadvantaged, non-home- less people seems relevant. Conclusions for policy and practice The successful Action Plan for Social Support of over a decade ago may explain the decline in the proportion of infectious diseases and traumata and the shifts within the main category of psychiatric diagnoses. Improved healthcare and support, alongside housing assistance,