From: Cameron Mustard
The second issue is much more of an obstacle. The original methods of the deliverable used the consolidated treatment information from MHSIP and MHMI databases to identify who in the population was in treatment for mental health disorder. Each person in treatment was then classified to one of three hierarchical categories of disorder (major, minor and other) on the basis of the diagnostic information in the two databases. For the observation period of interest to the deliverable, the DX information in MHMI was reasonably good, if one looked in the correct fields (don't use the case 'opening' DX, which is just white noise... one must use the case close DX.... which may be 36, 48, 60 months into the future...) However, since then, the mental health division has lost its commitment to this resource, and there is clear evidence that the field sites are not entering the DX information consistently. More to the point, as Renee Robinson showed in her master's thesis, the collapse of quality of the information in the system is regional specific. Some regional managers have <apparently> told staff to stop doing data entry... others have continued to require the information. The result is that the population view is now badly corrupted within MHMI.