Design
Review
Answers are grounded in this paper and cite their sources.
Shifting focus to theatre efficiency over drug selection may optimize savings; leaves open validation via prospective economic trials.
Many National Health Service providers have entered an era of strict budgetary control, service cutbacks and bed shortages, in the face of increasing demand for services [1]. Drug budgets are an easily identifiable area for managers to make savings, as the costs are easily determined and there are often cheaper alternative drugs [2–4]. Anaesthetists are particularly likely to be affected by restrictions in drug budgets because of the nature of their speciality. The use of low-flow anaesthesia has been advocated [5–8] and some anaesthetic departments have imposed anaesthetic practice guidelines to reduce costs [9]. The use of nonmedical personnel to deliver anaesthesia has been advocated on grounds of cost [10, 11] and the economic consequences of epidural analgesia in childbirth have also been subject to scrutiny [12]. Some workers [13, 14] have advocated the use of volatile anaesthesia rather than propofol for day-case anaesthesia in spite of an increase in postoperative nausea. This trend is worrying some observers [15] especially as differences in costs are often measured in pence per minute without regard to quality or cost-effectiveness analysis [4, 16]. I have sought to answer some of the economic questions surrounding the cost of anaesthesia, based on published evidence and internal audit where external data are not available. In particular the issues surrounding the cost of intravenous anaesthesia are examined as this is a common area where cost is used as a reason for restricting anaesthetists' choice of technique [17]. There are four commonly used methods for economic analysis in health care [16] which differ in the types of input measurement (resources consumed) and output measurement (health improvement) used: 1 cost-minimisation analysis; 2 cost-benefit analysis; 3 cost-effectiveness analysis; 4 cost-utility analysis. These are summarised in Table 1. There are also various definitions of types of cost [18], which need to be remembered when discussing economic data. Some of these are illustrated in Table 2. Drummond [19] has pointed out ‘methodologic weaknesses’ and the ‘limited availability of credible studies’ in the field of pharmacoeconomics. From an examination of published economic data regarding anaesthesia it is clear that most workers have concentrated on direct cost comparison or cost-minimisation analysis without taking into account relative cost benefits of different drugs. There is clearly a need for more work to be done using more appropriate analyses [20–24]. There are few published data regarding the precise cost of anaesthetic drugs used in particular procedures. A detailed survey in the Norfolk and Norwich Day Procedures Unit in 1994 [25] documented all drugs, disposables, time spent in different areas and staff present, for 530 operations over a 28-day period. Costs were categorised to anaesthetic drugs and disposables, surgical drugs and disposables, operating theatre staff costs, ward staff costs and overhead costs. The results were grouped by procedure code and summarised using a relational database. The anaesthetic drug costs for the five most common procedures are shown in Table 3. The range of cost for the general anaesthetic component was £6.00 to £8.08, although the mean duration of surgery for these procedures ranged from 6.6 to 31.9 min. This illustrates how most of the anaesthetic cost is incurred during induction of anaesthesia and that it costs relatively little to maintain anaesthesia (Table 4). Even with intravenous anaesthesia the cost of maintaining anaesthesia is small (approximately 13 pence per minute for a 70-kg patient) and compares well with the cost of a telephone call. When other costs of surgery were included (surgical costs, theatre staff, ward staff and overhead costs) it was clear that for short day-case procedures the anaesthetic drugs accounted for less than 4% of the total cost of the procedure. Churnside et al. studied in-patients as well as day cases in a similar survey [26]. They calculated the anaesthetic drug costs at less than 1% of the total cost for major abdominal surgery. To investigate the costs of different types of anaesthetic I studied 21 laparoscopic sterilisation procedures in more detail (Table 5). Five anaesthetists used three different anaesthetic techniques. The range of anaesthetic costs was small (£12.43 to £15.30) and there was no correlation between anaesthetic type and anaesthetic cost for these short procedures. Differences arose largely as a result of other drugs given such as neuromuscular blocking drugs and reversal agents. When all costs were included it was interesting to note that cases receiving the cheapest anaesthetic ended up with the highest total cost. My data are supported by the findings of Bailey et al. [7], who also showed that drug costs in 1993 were actually equivalent to those in 1959 when inflation was taken into account. Sensible managers should concentrate efforts on reducing the major cost components (staff, overheads and surgical disposables) and other important factors (e.g. admissions following day-case surgery, patients failing to arrive for surgery, late cancellations and operating department efficiency) before attempting to restrict the anaesthetist's choice of anaesthetic. Unplanned admission from the day surgery unit for an anaesthetic-related reason is a cause for concern. I have estimated that the overnight ‘hotel’ cost of an acute bed in our hospital is approximately £265. There is also the cost of nursing these patients, any treatment they might require and potentially lost revenue for the host unit if these admissions prevent routine admissions for elective surgery, which is particularly likely during the winter months. Seventy per cent (56 patients) of the admissions from our day procedures unit in 1993 were due to nausea and vomiting alone. Therefore it would seem sensible to use anaesthetic drugs which are known to have a low associated incidence of nausea and vomiting [27], even though their cost may be slightly higher. Raftery & Sherry [28] showed that 5% of patients who received propofol anaesthesia during assisted conception therapy were admitted overnight compared with 21% of patients receiving enflurane anaesthesia (p < 0.05). There were also significant differences in nausea and vomiting between these groups. The intelligent use of information technology should enable units to improve their utilisation of operating theatre time and has been shown to assist in anaesthesia cost containment [29]. By loading lists efficiently to avoid under or over utilisation it should be possible to treat the maximum number of patients in the time available and minimise variable and semifixed costs. A 30-min delay to the start of an operating session has far more economic impact than the choice of anaesthetic [30]. Most operating departments have a single postanaesthetic care unit, which may receive patients at random intervals from operating theatres. The rate of discharge of patients from these units can directly affect the number of patients it is possible to treat and so affect overall costs [31–33]. Therefore it would be sensible to encourage the use of anaesthetic drugs and techniques which result in rapid recovery from anaesthesia and a low incidence of side-effects such as nausea and vomiting, to minimise the risk of overloading the recovery unit. Although these drugs may be more expensive, it may be more cost-beneficial to use them overall [34–36]. Further studies, especially from UK institutions, are needed to evaluate this because recovery and turn around time in the UK is a particularly significant factor for most of us. Some anaesthetic techniques take less time than others. In my survey I found that the two anaesthetists who used intravenous anaesthesia and spontaneous ventilation took an average of 2 min less per case (Table 5) than their colleagues who intubated their patients and used enflurane maintenance. Over the course of an operating session of 10 cases this would result in a saving of 20 min and would enable another laparoscopic procedure to be performed. If this theoretical advantage were turned into a real increase in throughput, it would result in significant cost-benefits to the unit. Failure to arrive for surgery is a problem for many surgical units in the UK [37]. One centre has reported a ‘Did not attend’ rate of 36% for a day-case surgery list [38]. I have estimated the average cost per theatre per annum in the UK to be approximately £50 000–£75 000. Late cancellations can also be a cause for operating theatre under-utilisation [39]. Good management should be directed towards minimising the risk of this happening, by the use of patient reminders, mutually convenient admission dates, pre-admission assessment clinics and short notice admission lists. The economic benefits achievable here are potentially much greater than any possible savings from cost-minimisation of anaesthetic drugs, although they may be more difficult to achieve and quantify. It has been estimated that each treatment episode ‘saved’ in a day procedures unit is worth approximately £320 to the hospital's income [37]. A well-designed management system would save a typical two-theatre unit approximately £64 000 per annum if only four cases out of a total throughput of 160 cases per week are ‘saved’. Early return to work represents an important cost-benefit, and can lead to surprising conclusions. Heikkinen et al. [40] showed that laparoscopic hernia repair led to an earlier return to work than Lichtenstein hernia repair (median (range): 14 (8–26) days compared to 19 (5–40) days). Although the laparoscopic repair was a more expensive procedure ($1395 compared to $878) the median total cost was less, due to the earlier return to work ($4796 compared to $5320). Again there are very few published data regarding the effect of anaesthesia on time off work. Enlund [41] showed that patients returned to work 1 day earlier (comparing medians) when they had received propofol/alfentanil anaesthesia compared to thiopentone/nitrous oxide anaesthesia for a minor day-case procedure. The increased cost of anaesthesia was more than compensated for by an earlier return to work with a net gain of 111SEK (£8.74) per patient to the social insurance system. The overall benefit of the propofol/alfentanil combination corresponded to almost twice the increased cost of anaesthesia. There would, however, need to be a willingness to be prepared to pay extra for treatment options and technologies [42] which do not directly benefit the hospital, but benefit the State and self-employed patients rather more. More recently, Sinclair et al. have shown that optimisation of intravascular volume during surgery shortened the length of hospital stay by 39% in a high-risk group of patients [43]. So it would appear that clinical management during surgery and anaesthesia might affect the postoperative course for some patients. There is very little published work on the effects of anaesthesia on postoperative stay in hospital, but as the current trend is to shorten postoperative stay for many procedures, the effects of anaesthesia will become increasingly significant. I have shown that for short procedures the differences in anaesthetic costs are small and do not necessarily depend on the choice of anaesthetic agent. Also the anaesthetic drugs account for a small percentage of the overall cost of surgery. Comparatively small savings can be made during long operations using volatile agents provided that low gas flow systems are used effectively [44]. However it must be remembered that in a typical general hospital most operations take less than 60 min (Fig. 1) so the savings are not great in many cases. Frequency histogram of 16 384 general anaesthetic cases performed at the Norfolk and Norwich Hospital in 1996. Source: Norfolk and Norwich Hospital ORSOS theatre information system. Median = 45 min, mode = 35 min. The use of propofol anaesthesia in particular has been criticised for being expensive. However the costs are small and can be reduced significantly by the use of adjuvant drugs such as midazolam [45–47] and remifentanil [48]. The benefits of this type of anaesthesia may outweigh the extra cost, by increasing the chances of a speedy and complication-free recovery. A large multicentre trial [49] showed that there was significantly quicker recovery and a lower incidence of nausea in the first 2 h following major abdominal surgery when intravenous anaesthesia was used. These characteristics would benefit the in-patient operating theatre block as much as the day surgery unit in terms of increased overall efficiency and throughput. More data, using relevant pharmacoeconomic analyses, are needed. If anaesthetic departments act purely with regard to cost-minimisation analyses they assume that clinical differences between anaesthetic drugs are insignificant and they may as well use drugs which have been available since the late 1950s in order to minimise costs completely. Anaesthetic cost minimisation may end up costing the unit more, elsewhere in the system [50]. If new anaesthetic drugs and technologies are ignored on grounds of cost, then future investment in anaesthesia research and development by industry is likely to decline [51]. Far more significant savings could be made if operating units optimised theatre efficiency and introduced patient management systems to reduce late cancellations and nonattendance. Finally, patients' perceptions of the quality of anaesthesia are often ignored. Orkin [52] found that 72% of patients were most worried about postoperative emesis before their operation compared with 8.5% who were most concerned about postoperative pain. His paper suggests that most patients would put a higher value on the prevention of PONV than many anaesthetists, who seem prepared to tolerate an increased incidence of PONV [13] in order to save a dollar or two.
No takes yet. Share an insight, caveat, or question.
W. L. Rowe (1998) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: