Introduction

There are many factors to consider when analyzing readmissions. For example:


• What cases, if any, should be excluded prior to defining index cases? (e.g. non-inpatient hospitalizations)
  
• Who is the cohort for whom readmissions are to be counted?
  
• Which hospitalizations should be excluded from being counted as readmissions (e.g., transfers, concurrent stays, deaths, elective, extended care)?
• How does one define such exclusions?

Approaches

Different approaches have been adopted at MCHP depending on analysis needs:


• Project-specific programs have been developed for MCHP researchers (see example below).
  
• The COMBINE macro has been used to generate readmission information (e.g. for SBGH data).
  
• A user-friendly graphical user interface has been developed. Some work has been done on this approach by Charles Burchill.

EXAMPLE: Readmissions were run for urban hospitals for 14 medical/surgical diagnoses for a range of 6 years (1989/90 -1994/95) of Hospital Discharge Abstracts data. The ADRG's were used to select these diagnoses:

89 - Simple Pneumonia 96 - Bronchitis and Asthma 121 - AMI 127 - Heart failure 148 - Major Bowel 157 - Anal / Stomal 161 - Inguinal / Hernia 182 - Digestive disorders 197 - Open Cholecystectomy 336 - Prostatectomy 358 - Uterine / Adnexal 370 - C section 373 - Vaginal delivery 430 - Psychoses 493 - Lap. Cholecystectomy

The readmission for any of the 14 diagnoses was defined as an inpatient admission to any hospital for any reason within 30 days from the index admission (from date of discharge). The index admission was restricted to:


• Winnipeg Urban Hospitals
  
• a LOS restriction of 30, 60 and 90 days depending on the diagnosis.
  
• In another scenario First Nations, residents from the core area, and the high severity cases (using RDRG's®) were removed.

Different index admission restrictions can be defined and re-programmed easily.