A Markov decision model was used to synthesise the data to evaluate the cost-effectiveness of interventions to treat urinary tract infections. The time frame was 30 years, to consider the more serious conditions that can follow on from recurrent urinary tract infections, such as renal scarring, chronic kidney disease, and renal failure. The authors stated that the perspective was that of the health care provider. Boys and girls were evaluated separately due to differences in some of the probabilities for the model.
The effectiveness data that produced the transition probabilities for the Markov model were identified by a systematic review of relevant literature. The search strategy and the inclusion criteria were reported. Published evidence and North American, Swedish, and English guidelines on urinary tract infections were used for the components of maximum care. A published study in Dutch primary health care and Dutch primary care guidelines were used for usual care and these guidelines were used for the improved usual care.
Monetary benefit and utility valuations:
Quality-of-life weights, based on patient preferences, were identified from published studies. For chronic kidney disease, the weight was measured using the standard gamble method and, for terminal renal failure and after kidney transplant, the time trade-off method was used.
Measure of benefit:
The measure of benefit was quality-adjusted life-years (QALYs), which were discounted at 4% per annum.
The cost categories were general practitioner visits, paediatrician or nephrologist visits, diagnostic tests, and antibiotic medications. These costs were based on the guideline prices for the Netherlands. They were reported in Euros (EUR) and were discounted at 4% per annum. The price year was 2006.
Analysis of uncertainty:
Probabilistic sensitivity analysis was conducted to assess the uncertainty, using first- and second-order Monte Carlo simulations. One-way sensitivity analysis was also performed to examine if the model was robust to changes in the discount rate.