Research Article

Extending the use of clinical audit data: a resource planning model.

BMJ 1990; 301 doi: https://doi.org/10.1136/bmj.301.6744.159 (Published 21 July 1990) Cite this as: BMJ 1990;301:159
  1. B W Ellis,
  2. R C Rivett,
  3. H A Dudley
  1. Ashford Hospital, Middlesex.

    Abstract

    OBJECTIVE--To create a means by which we can examine and understand the interrelations among the fundamental elements of hospital inpatient care (patients, beds, theatre time, and staff). DESIGN--Predictive study of resource utilisation based on a computerised clinical information system of five years' audit data from a surgical management system. SETTING--One surgical firm (of one consultant, one registrar, and one preregistration houseman) in a district general hospital. PATIENTS--5267 Patients whose admission records were part of the five years' audit of surgical management. MAIN OUTCOME MEASURES--Mean length of stay; number of occupied beds; turnover interval; throughput (patients/bed); percentage elective theatre occupancy; waiting time for elective admissions; and theatre, hotel, and total costs. RESULTS--Predicted outcome was analysed in the model, taking the actual outcomes in 1988-9 as baseline values, for four clinical scenarios: an increase in accident and emergency admissions, a reduction in beds, a reduced length of stay, and creation of a new firm. Baseline values showed a mean stay of just over five days in 15 beds and with a theatre occupancy of 94%; the total cost was 812,000 pounds (hotel costs 597,000 pounds). Increasing the accident and emergency admissions to 460/year (19%), based on projected trends from 1984 to 1988, resulted in increased hotel costs (55,000 pounds) and reducing bed numbers (by halving admissions) in decreased use of theatres to 71%, decreased throughput, and increased waiting time, from 20 to 92 weeks, at a saving of 99,000 pounds (12%). Reducing stay marginally reduced bed occupancy (8%) and hotel costs (14%), and creating a new surgical team considerably reduced bed occupancy (14%) and waiting time for elective operations (by 20%). The minimum number of beds for referrals, accident and emergency admissions, and planned admissions was 9.0; that for urgent elective admissions was 3.3 and for non-urgent admissions was 2.4. CONCLUSION--A well designed clinical information system with the routine collection of data can provide the necessary output data to enable resource modelling. IMPLICATION--Use of such a model will allow clinicians to participate in resource planning on the basis of what is actually happening within the hospital.