Proefschrift

CHAPTER 3.3 158 METHODS Data sources We used hospital discharge data from the Health Informatics Centre, Ministry of Health, Malaysia from January 1, 2007 to December 31, 2016. Briefly, the Hospital Discharge Register represents a module within a centralised database known as the Patient Management Information System ( Sistem Maklumat Rawatan Pesakit ). It has been in operation since 1999 with the compilation of aggregated data on inpatient admissions. Then, granular patient records for Ministry of Health (MOH) hospitals became available by gradual increase in hospital participation over the years. Based on data from the Health Informatics Centre, inpatient admissions to public (MOH) hospitals make up to about 83% of HF hospitalisations in the country between 2008 and 2016, with the rest being hospitalisations in private, university and Ministry of Defence hospitals. We included only data for MOH hospitals to maintain uniformity as data from other hospital types were available for only a later portion of the study. Mortality outcomes were determined via linkage with the National Mortality Database from the National Registration Department and Department of Statistics by matching on the national identification number or passport number for foreign nationalities in combination with date of birth and sex.6 Overall, death registration in Malaysia has a coverage of more than 90%; there is complete registration in West Malaysia but under-reporting remains in East Malaysia (comprising the states of Sabah and Sarawak).6–8 Deaths in hospitals were medically-certified by attending physicians or coroners. On the other hand, deaths in the community were certified by informants such as the policemen and medical assistants and these were considered non-medically certified.9,10 For the reporting of all-cause mortality, we considered all deaths, regardless of whether they were certified by medical personnel whereas for the calculation of cause-specific deaths, only those which were medically-certified were analysed. Study population Patients who were aged 20 years and above, who were admitted to a MOH hospital for an incident HF hospitalisation were included. A discharge diagnosis of HF was defined by International Classification for Diseases version 10 (ICD-10) code I50 while

RkJQdWJsaXNoZXIy MjY0ODMw