Abstract
Background
Balloon kyphoplasty (BKP) is a minimally invasive cementing procedure, occasionally used in patients with painful vertebral compression fractures (VCF). In this multicenter Swedish RCT, we evaluated the cost-effectiveness of BKP compared with standard medical treatment, Control, in osteoporotic patients with acute/sub-acute VCF (<3 months). In a multicenter European clinical study (FREE trial) including 300 patients and FU after one year, BKP was suggested to be a safe and effective procedure in selected patients. The current study includes the Swedish patients in the FREE trial Method: Hospitalized patients with a back pain level of at least 4/10 on a visual analogue scale due to of VCF between Th5–L5 (confirmed by MRI) were randomized to either BKP or Control treatment (standard medical treatment with pain medication and functional support). All VCF-associated costs (hospital, primary care, rehabilitation, community care, private care, pharmaceuticals, assistance by relatives, work absenteeism) were identified and reported from the perspectives of cost to society, and costs to the healthcare system. Primary outcome was quality of life change (QoL) measured with the preference based EQ-5D instrument. The accumulated quality adjusted life years gained (QALYs) and costs per QALY gained was assessed. Willingness to pay (WTP) for a QALY gained in Sweden was estimated at approximately SEK 600,000 (EURO 62,500). Sensitivity analyses were performed.
Results
Between February 2003 and December 2005, 70 patients were randomized to BKP (n=35) or to standard medical treatment (n=35). Three patients in the Control group declined to participate in an economic evaluation, and only patients answering EQ-5D at all FU occasions (1-3-6-12-24 months) were included in the analyses, leaving in all 63 patients, BKP=32, Control=31. Baseline data were similar. The mean age in the BKP group was 72 years (71% women) vs. 75 years (78% women) in the Control group. Baseline difference in QoL was adjusted for using statistical methods. There were no cross overs. Four patients in the BKP group and three patients in the Control group died within two years of causes not related to the VCF. Costs were collected using “cost diaries” in mailed questionnaires after 1-3-6-12-18-24 months. Costs and EQ-5D values (0 at FU after death) were carried forward. Total mean societal cost per patient for BKP and Control was SEK 160,017 (SD 151,083) and SEK 84,816 (SD 40,954), respectively. The difference was significant 75,198 (95% CI 16,037 to 120,104). The accumulated mean difference in QALYs was 0.085 (−0.132 to 0.306) units in favour of BKP. Cost per QALY gained using BKP was SEK 884,682 (EURO 92,154) with high uncertainty assessed using the bootstrapping technique, and demonstrated on the cost-effectiveness plane and on the acceptability curve. When the EQ-5D values from all patients in the FREE trials were included in a sensitivity analysis, cost/QALY was SEK 359,146 (EURO 37,411) Conclusion: Costs were significantly higher for BKP compared with standard medical treatment, with no significant difference in QALYs gained. In this selected patient population with vertebral compression fracture due to osteoporosis, BKP could not be concluded as cost-effective after two years
