Development of a Predictive Computer Model for Assessment of Community Health from an Annual Health Check-up

Abstract

Health check-up  program  has been widely  used to assess community health status and to identify  people who are at risk according to a significant range of laboratory parameters. The laboratory results from a single  health check-up  is, however, a snapshot of individual  health at a certain  time point in which fluctuation due  to biological variation  resulting  in the unreliable measurement of the community health.  Laboratory result  data collected  from  multiple  health check-ups could  be used to improve  the assessment of community  health.  However,  it is not practical  to perform frequent health check-ups  due to limited resources such as workforce, time and  budget.  In  this study,  the  Markov  Chain  computational model  was developed  to predict the average  distribution of health  parameters using  data from  a single  check-up. The  results demonstrated that the established  model gave a better prediction  of the average distribution from three  consecutive  check-ups than a single check-up. Therefore, this could  be used by Medical Technologists and  healthcare  promoters to  predict   health  status   to  gain   better   insight  of community health. This study also revealed  the innovative  use of longitudinal health  check-up data in health promotion,