The use of micro-costing in economic analyses of surgical interventions: a systematic review

Accurate information regarding the cost-effectiveness of existing and novel surgical interventions is vital to inform policy and ensure scarce resources are used to optimal benefit. Central to the meaningful assessment of cost-effectiveness is the accurate evaluation of resource use [1]. This is commonly undertaken using a ‘top-down’ or ‘gross-costing’ approach whereby average costs are used to estimate resource use. For instance, in England, national reference costs based on healthcare resource groups (HRGs) can be used to estimate the average cost per inpatient episode for groups of surgical procedures (e.g. “very major hip procedures for non-trauma”) [2]. While this method is straightforward, it is too crude for many purposes. For example, it is not possible to compare the costs of two different surgical procedures within the same procedure group (e.g. cemented vs hybrid total hip replacements) or evaluate a modification to an existing procedure when the procedure itself is unchanged (e.g. single port vs standard multiple port laparoscopic appendicectomy). Furthermore, gross costing approaches may not be suitable for evaluating novel surgical interventions (e.g. robotic cardiac surgery) which do not fall within existing HRGs. Micro-costing is a method that allows for more precise assessment of the economic costs of a healthcare intervention [3]. Defined as the ‘direct enumeration and costing of every input consumed in the treatment of a particular patient’ [4], micro-costing attempts to measure costs of a service as accurately as possible. The process has three stages; the identification of all resources involved in the provision of care (e.g. human-resources/theatre-time/consumables); accurate measurement of each resource (for example using time-and-motion studies); and valuation of the resources used. Although this approach is time-consuming, it may more accurately reflect the cost of surgical procedures, especially if the procedure is new or includes the use of expensive implants or consumables. It may therefore be the most appropriate method of resource use assessment alongside surgical trials.

The potential value of this methodology has already been identified and a framework has been developed for applying micro-costing methodology to cost evaluations of surgical technologies [5]. This framework focuses specifically on direct costs which reflect the price of resources directly attributable to the procedure rather than indirect costs (e.g. overheads) which have to be estimated using an allocation formula [5]. The framework divides the direct costs into two categories; fixed costs which do not vary according to the level of activity (e.g. personnel costs and medical devices e.g. robotic systems) and variable costs such as reusable instruments and disposables which vary according to the type or number of procedures performed. A recent review of use of micro-costing in bariatric surgery suggests that this framework may be more widely applicable to surgical interventions [6], but further work is required to determine whether it is generalisable to a range of surgical procedures.

Despite the potential benefits of micro-costing, it is time and resource intensive and this may limit its utility in clinical trials. Work is therefore needed to explore whether a more targeted approach may be possible. It is hypothesised that surgical interventions will have ‘cost-drivers’ which create or drive the majority of the cost of the activity. If these are consistent across surgical interventions or can be easily identified, it may be possible to develop a simplified micro-costing framework which could be used in future surgical trials.

The aim of this systematic review was therefore to identify and critically appraise published studies reporting the use of micro-costing in economic analyses of surgical interventions to explore the resources included in each study; the methods used for measuring and valuing resource use and reported cost-drivers of surgical interventions to inform recommendations for the future use of micro-costing in surgical interventions.

The protocol for this systematic review was registered on the PROSPERO International prospective register of systematic reviews (reference number CRD42018099604) before data extraction commenced.

Micro-costing may offer a more accurate method for costing in economic analyses of surgical interventions, but this review suggests that current use of the methodology is inconsistent and, in many instances, lacking in methodological rigor. There is a lack of consistency regarding terminology used; the types of resources included and the methods by which these are measured and valued. Almost half of studies used retrospectively collected data (e.g. from hospital administrative or information systems). Although routinely available data sources may be accurate, especially in modern ‘fee-for-service’ settings, other reported methods for resource use identification such as case-note review or operating room logs may be less robust [13‐15], potentially compromising the validity of the micro-costing approach. Over three quarters of all studies were undertaken at a single centre limiting the generalizability of the results. Almost half of the included studies did not report resource utilization and unit cost data separately. Instead summary costs were presented with limited disaggregation of individual costs or transparency regarding precisely which resources had been included making meaningful interpretation difficult. Some good examples of micro-costing surgical interventions, however, were identified [16, 17] and have subsequently been published [18].

There was significant variation in which aspects of the patient pathway researchers elected to micro-cost. Most studies costed both the surgical procedure and the associated inpatient stay, but others also included pre-operative planning/investigations, complications and follow-up required following discharge. To an extent, this variation may appropriately reflect the diversity in the aims of micro-costing studies. For economic evaluations comparing two different surgical interventions researchers may legitimately focus on the incremental cost differences between the two procedures and exclude costs that are common to both procedures (e.g. pre-operative visits). Furthermore, for economic evaluation, methodological guidelines emphasise that the level of resource use detail required is lower for items with a small relative impact on total costs (e.g. volatile anaesthetic agents) [19]. In contrast, for studies aiming to compare the actual costs of surgery to “top down” costs (e.g. hospital reimbursement), it is important to include all elements contributing to cost. The level of detail may also be constrained by study setting and in particular the granularity of the unit costs available for valuing resources. Studies in settings with itemised fee for service hospital reimbursement have greater scope to cost resource use in much finer detail. Approximately one in four studies compared a measured cost of performing a procedure, as determined by micro-costing, with procedure reimbursement rates, which were often considered to be insufficient. Many of these studies were undertaken by clinicians. The scope of micro-costing (e.g. procedure based or full admission) and the economic perspective (hospital or wider health service) in such studies should be rigorously appraised to ensure it is not biased by clinicians’ financial interests. Cost-drivers were identified and reported in most studies, but these differed according to which aspects of the patient pathway were included. There is a need for a standardized definition of a cost driver in the literature so that authors are consistent with this terminology and how it is measured and reported. Although, it was clear from the reported studies in this review that the cost drivers were the resources that involved the highest proportion of the total costs of the surgical pathway and/or surgical procedure, this proportion may in some studies have been below 50% of total cost and/or may have subsumed several cost drivers instead of only one cost driver per pathway or procedure.

This review has demonstrated the need for specific guidance for researchers undertaking micro-costing. Work to develop a checklist for the conduct, reporting and appraisal of micro-costing studies in healthcare is ongoing [20]. This checklist may improve the quality of future studies but is not specific to surgery which presents unique challenges. A framework for costing surgical technologies [5] provides formulae for estimating the fixed (device and personnel) and variable (re-usable equipment and disposables) costs of surgery. However, the framework only considers the costs of the operative procedure which limits its applicability for trials of surgical procedures likely to have resource-use implications beyond the operating theatre such as differential length of inpatient stay; use of intensive care or need for follow-up investigations. The framework also excludes indirect costs (overheads) of surgery which are necessary for micro-costing analyses wishing to compare surgical care with non-surgical care or compare the cost of surgery with procedure reimbursement values. Further work is therefore needed to develop more simplified recommendations for the use of micro-costing in surgery generally and specifically for efficient and effective use of the methodology in surgical trials. Recommendations based on the findings of this review are summarised in Table 5.

This is, to our knowledge, the first study to systematically identify and critically-appraise the use of micro-costing as a method of resource-use assessment in surgery. We used published search strategies for micro-costing [6, 7], however due to the lack of standardisation in micro-costing terminology and excluding non-English language papers, our search may have overlooked other relevant studies. The review was restricted to studies meeting our prespecified definition of being a ‘sufficiently-detailed micro-costing exercise’. This led to the exclusion of a large number of studies at the full-text screening stage. Inclusion of these studies would not have been informative, but it is important to note that costing studies in surgery more generally may be methodologically less robust than those included in this review. The majority of included studies scored highly on the CHEC checklist despite an objective lack of methodological rigor. This highlights the need for specific recommendations for the design, conduct and critical appraisal of micro-costing studies to improve the quality and value of this work.

Micro-costing produces an accurate assessment of resource-use but the methodology is time consuming and resource intensive and universal application alongside surgical trials would not be practical [3, 21]. A more targeted or hybrid micro-costing approach, however, may have value [22]. Almost all studies included in the review identified one or more ‘cost-drivers’ that represented the greatest proportion of the costs for their procedures. These differed according to the aspects of the patient pathway considered as for example, operating room costs were the main cost-driver in studies just costing the surgical procedure but were less important in studies costing the full episode of care. Cost-drivers may also be influenced by other factors such as the specific research question and the intervention under study. For example, a study comparing the cost of two surgical procedures identical except for the implant used might reasonably focus on implant cost and detailed observations of surgeon time spent inserting the implant. Focusing attention on identifying and accurately costing these key elements of care while applying less detailed methods (e.g. record review or expert opinion) for other components may allow the benefits of micro-costing to be realized in an efficient way. Process mapping with experts including surgeons, nurses, other allied healthcare professionals and if appropriate, patients themselves may be an effective way to identify key cost elements. More targeted, efficient micro-costing of key cost-drivers may also allow this methodology to be applied over a wider number of centres in a trial setting improving both the accuracy of the costing data obtained and the generalizability of the results [23]. Further work, however, is needed to determine if this targeted approach would be feasible alongside a surgical trial.

As health systems and technologies become more advanced, routinely-available data may have increasing utility for resource-use assessment and may make future large-scale micro-costing studies possible. Healthcare provider cost-accounting systems that capture details on all resources used in an episode of care combined with electronic theatre systems that could provide proxy time and motion data may represent a time and cost-efficient method for detailed resource-use assessment. Such systems and methodology, however, would need to be consistent across centres if the costs generated were to be comparable. This may be challenging, especially across different healthcare settings and geographical locations.

Micro-costing has the potential to improve the accuracy of economic analyses of surgical interventions by providing more accurate assessments of resource-use but the overall quality of existing studies is poor. There is a need to improve the consistency and efficiency of micro-costing in surgery if the potential value of the methodology is to be realised. This review highlights a number of ways in which this could be achieved. Using standardised terminology for micro-costing studies; focusing on identifying and accurately costing cost-drivers relevant to the specific research question and transparently reporting disaggregated costs for each included resource may represent a simple strategy for improving the design and delivery of future studies.

Micro-costing may provide a more accurate method of resource-use assessment in economic analyses of surgical interventions and could improve the value of economic evaluations conducted alongside surgical trials, but this systematic review suggests that current use of micro-costing in surgery is inconsistent and lacking in methodological rigor. Using standardised terminology for micro-costing studies; focusing on identifying and accurately costing cost-drivers relevant to the specific research question and transparently reporting disaggregated costs for each included resource may represent a simple strategy for the optimal design and delivery of future studies.