Analytical approach:
A 15-year Markov cost-utility model was developed for the ongoing risk of hip-replacement failure and the need for revision. The model had one-year cycles. After replacement surgery, patients could have two revision surgeries. Resurfacing patients could have a revision or a replacement; after a revision, they could have a replacement, and after replacement, they could have two revision surgeries. The authors stated that they took a health care system perspective.
Effectiveness data:
The primary effectiveness data were the time to replacement failure and the need for revision for each type of replacement. These were from the National Joint Replacement Registry of the Australian Orthopaedic Association (184,629 patients). Estimates were calculated by age group and gender. Surgery-related mortality was from a Swedish registry, and complications were from the Hip Improvement Project (HIP), in Alberta.
Monetary benefit and utility valuations:
Utility scores were applied for the first year after surgery, for each following year, and for whether or not a complication was experienced during surgery. The utilities depended on whether the surgery was primary, revision, or conversion. They were from the HIP, which used the SF-36, which was converted to SF-6D, and analysed by age group (>55 or ≤55 years) and gender. A common ratio of utilities for patients receiving replacement or resurfacing was applied. A ratio based on the HIP was applied for revision utilities. The utilities for complications were derived from a published study; they were assumed to apply for three months.
Measure of benefit:
The primary measure of benefit was quality-adjusted life-years (QALYs). Benefits were discounted at 3% annually.
Cost data:
Various Canadian or Alberta-region unit cost sources were used. The data included physician care, surgery, prosthetics, direct and indirect hospital costs, revisions, and care after surgery. Surgical costs included transfusions, cement, surgery time, and number of nurses and assistants, based on chart review. Revision and conversion costs were assumed to be identical, and were a proportion of primary surgery costs, based on a Canadian systematic review. Care after surgery included orthopaedic surgeon visits, physiotherapy, and prescription analgesics. Complications were assumed to cost the same regardless of surgery type, based on a publication. All costs were inflated to 2011 Canadian dollars (CAD) using the health care Consumer Price Index. Future costs were discounted at 3% annually.
Analysis of uncertainty:
Probabilistic sensitivity analyses were undertaken for the main analysis and seven subgroup analyses by gender and age group. The results were presented as cost-effectiveness scatter plots and cost-effectiveness acceptability curves. Deterministic threshold analyses were undertaken to assess the revision or conversion probability, and the cost of resurfacing, at which replacement was no longer less effective and more costly.