D-6 OPTIMIZING THE NUMBER OF NEONATAL INTENSIVE CARE UNIT BEDS IN BRITISH COLUMBIA

Monday, October 25, 2010: 5:45 PM
Grand Ballroom East (Sheraton Centre Toronto Hotel)
Derrick L. Fournier, PEng., MBA and Gregory S. Zaric, PhD, The University of Western Ontario, London, ON, Canada

Purpose: Each year a small number of expectant mothers with high-risk pregnancies in British Columbia (BC) are sent to the United States (US) because of a lack of neonatal intensive care unit (NICU) beds. We sought to determine the impact of changing the number of NICU beds in BC on the probability of transfer to the US and on overall system costs.

Method: We developed a discrete event simulation model to determine the probability of transfer to the US based on bed availability at each hospital in BC.  We modeled four types of NICU beds (Level I, IIA, IIB and III) which were differentiated by their ability to handle different levels of patient acuity. We used a Markov chain to model daily changes in a neonate’s health status and assumed that inter-hospital transfers were based on geographic distance. We obtained data on the current number of beds of each type from the BC Health System, and we obtained data on birth rates, length of stay and costs from the Canadian Institute for Health Information and published reports. We varied the number of beds of each type at each hospital and observed the impact on total costs and the probability that a baby would be transferred to the US.

Result: Adding a total of 13 Level II beds province-wide caused yearly system costs to be reduced by $803,000.  In addition, the combined average Level II and Level III bed utilization was reduced from 88% to 83% and the probability of transfer to the US was reduced from 0.018% to 0.0074%.  Adding 3 Level III beds caused system costs to be reduced by $278,000, reduced average Level III bed utilization from 87% to 84%, and reduced the probability of US treatment from 0.018% to 0.012%.  Adding Level I beds increased system costs but had no effect on the probability of transfer to the US. 

Conclusion: Adding a small number of Level II or Level III beds results in a reduction in the probability of transfer to the US. It also results in cost savings because of the significant cost of treatment in the US relative to Canada. We believe that the BC Health System should consider a modest increase in province-wide NICU capacity.

Candidate for the Lee B. Lusted Student Prize Competition