Health care systems frequently have to decide whether to implement interventions designed to reduce gaps in the quality of care. A lack of information on the cost of these interventions is often cited as a barrier to implementation. In this article, we describe methods for estimating the cost of implementing a complex intervention. We review methods related to the direct measurement of labor, supplies and space, information technology, and research costs. We also discuss several issues that affect cost estimates in implementation studies, including factor prices, fidelity, efficiency and scale of production, distribution, and sunk costs. We examine case studies for stroke and depression, where evidence-based treatments exist and yet gaps in the quality of care remain. Understanding the costs for implementing strategies to reduce these gaps and measuring them consistently will better inform decision makers about an intervention’s likely effect on their budget and the expected costs to implement new interventions.

Recent initiatives by the Department of Health to reform the health and therapy services place great emphasis on quality assurance, highlighting the need to evaluate performance by means of clinical audit. Whilst audit is an acknowledged method of evaluating the accuracy of treatment records, and the efficiency and effectiveness of practice, it can have a broader function. This paper provides the results of three audits of physiotherapy patient records, designed to determine ( 1) the range and extent of measurement used for objective phenomena, such as joint range and muscle strength, ( 2) the range of techniques used in the management of low back pain ( LBP) and following total knee replacement ( TKR) surgery, and ( 3) the effectiveness of physiotherapy intervention in the treatment of LBP and following TKR. The audits were conducted in five hospitals in three counties of the north of England. A total of 1254 physiotherapy patient records from 1994 through 1996 met the criteria and were audited. These records were selected from outpatients, orthopaedic, and rheumatology departments. The results reveal that apart from the initial assessment, recorded in more than 86% of cases, other requirements of documentation were recorded less frequently. The type of measurements used lacked objectivity for most parameters including objective phenomena, and the range of parameters used to monitor patient outcomes was extremely limited, with functional assessment notably absent. Treatment techniques for LBP were mainly electrothermal, McKenzie, and exercise: Passive mobilising techniques were used less frequently. Following TKR the Cryocuff® featured in a large percentage of cases for the treatment of oedema and pain. Surprisingly, "straight leg raise" was frequently used to improve quadriceps strength.The audit results indicate a need to establish the reasons for the poor quality of recording amongst clinical physiotherapists, in order that the recommended Chartered Society of Physiotherapists standards can be achieved. Further avenues for research and audit are identified.