67. Surgery

Costs and Cost-Effectiveness of Interventions

In today's resource-constrained world, policy makers increasingly need to be aware of the value of selective health care interventions. Cost-effectiveness analysis is one method that links inputs (costs) with the resulting health care gains measured along a common metric, usually using DALYs.

Even though an extensive body of literature examines the cost-effectiveness of a range of nonsurgical interventions in developing countries (Jha, Bangoura, and Ranson 1998), the literature examining surgical interventions in these countries is more sparse. Moreover, most of the available studies examine surgical interventions for specific conditions (Marseille 1996; Singh, Garner, and Floyd 2000). A common criticism of such studies is that they do not fully capture the choices policy makers face in real life. For example, policy makers must often choose between allocating resources for constructing several community clinics or a single district hospital, both of which provide a mix of surgical and nonsurgical services. Generally, the surgical ward in a district hospital will provide care for a wide range of conditions, such as trauma, childbirth, and abdominal conditions. We assume that for policy makers, knowing the cost-effectiveness of the surgical service, ward, or clinic (as an intervention) is more useful than information about the cost-effectiveness of each condition-specific surgical intervention. Unfortunately, no literature exists that examines the cost-effectiveness of a surgical service or ward. This section attempts to fill that void with respect to district hospitals and community clinics but not in relation to tertiary-level hospitals, which vary in size, available resources, and role from region to region, making it difficult to describe the cost-effectiveness of a prototypical tertiary hospital.

Method for Estimating Costs and DALYs

On the basis of the resource requirements listed in table 67.4, we developed cost estimates for each of the six regions defined by the World Bank. Table 67.5 details the assumptions and table 67.6 provides the regional costs. We defined the standard hospital in such a way as to facilitate comparisons across regions, conceptualizing it as a 100-bed hospital with a male ward and a female ward; two operating rooms; a recovery room, an intensive care unit, or both; an x-ray unit and an ultrasound machine; and a laboratory that can carry out basic blood chemistry tests, examine urine, and cross-match blood. This hospital also has an on-site laundry and kitchen and two vehicles to serve as ambulances. The staff consists of 6 doctors (4 primary care physicians, 1 obstetrician and gynecologist, and 1 general surgeon); 20 nurses; 6 midwives; 2 physiotherapists; and 6 orderlies. The costs of an anesthetist and x-ray technician have been included in the operating costs of the operating rooms and radiology area, respectively. The model assumes that the hospital averages 80 percent occupancy and that two-thirds of inpatients will be surgical cases.¹

We defined a standard community clinic (see table 67.4) as a facility of 100 square meters serving a population of approximately 20,000, staffed by a nurse or nurse-substitute, a skilled birth attendant, and an orderly. Such a clinic treats approximately 4,000 surgical cases per year, with a surgical case being defined as treatment of bruises, simple cuts requiring suturing, foreign body removal, drainage of abscesses, basic burn treatment, normal deliveries, and simple trauma.

As far as possible, we used standardized regional cost estimates provided to the authors. When such information was unavailable, we used our consensus judgment. Given the wide variation in costs between and within regions, we conducted sensitivity analyses to capture the range of possible outcomes. When more than one source of cost estimates was available, the mean of the estimates for that region were used as the best estimate and a high-low range was noted. However, in many cases, only a single cost estimate could be obtained, in which case the data provide a point estimate,² and we vary the cost estimate by 20 percent to obtain a high-low range.

Our calculation of the number of DALYs averted was based on the work of McCord and Chowdhury (2003), who calculate the DALYs averted by a 50-bed hospital in Bangladesh, as described in box 67.1. We adjusted this figure to reflect the bed size of our standard district hospital. In the absence of region-specific data, we applied this figure to all six regions after making suitable adjustments. For the community clinic, we estimated that such a clinic averts approximately 200 DALYs per year as a result of surgical treatment, primarily from the incision and drainage of abscesses and the preliminary treatment of burns. Because these DALY estimates are based on a single source, we vary the estimate by 20 percent to obtain a high-low range and apply these estimates across the six regions.

Results

Figure 67.1 presents the results of the cost per DALY averted calculations for a district hospital and community clinic. The low estimate represents the scenario in which the costs are the lowest and the DALYs averted are the highest—that is, the best-case scenario. In a similar vein, the high estimate is the worst-case scenario: the costs are highest and the DALY averted is the lowest.
[Figure 67.1]

The best estimates for cost per surgical DALY averted at a community health center (panel a of figure 67.1) hover in a narrow range between US$212 and US$241. The cost per surgical DALY gained at a district hospital is cheapest for Sub-Saharan Africa at US$33 (range of US$19 to US$102) and most expensive for Latin America and the Caribbean at US$94 (range of US$47 to US$164).

Standard economic costs can differ from costs actually incurred in service delivery, both because in practice not all time may need to be paid for (for example, hospitals may be able to economize on staff because relatives help care for patients) and because low-cost solutions may be found (for example, use of paramedical staff members in place of professionals). Box 67.2 describes some of these strategies and compares the standard economic cost presented above with the much lower financial cost of a nongovernmental organization (NGO) hospital.

Discussion

The data in figure 67.1, when compared with similar data for other services presented in this book, indicate providing basic surgical services is relatively cost-effective. Figure 67.1 also indicates that, from a surgical perspective, the costs per DALY averted at a community clinic tend to be higher than those averted at a district hospital despite the lower costs of a community clinic. Although these observations may be taken as evidence that surgical services are best provided at the district hospital level, this goal may be impossible to put into practice. The type of surgical care provided at the community clinic level, though not resulting in a very large DALY gain, is nevertheless important. It is inconceivable to think of a community clinic that does not have facilities for minor foreign body removal, simple suturing of cuts and wounds, or splinting of simple fractures. Furthermore, community clinics' referral and primary treatment functions, which are hard to evaluate separately from the delivery of final treatment, are critical for many conditions, notably trauma.

Costs per surgical DALY averted at the district hospital level seem to fall into three groups. Sub-Saharan Africa and South Asia are the cheapest, with the best estimates of cost per surgical DALY averted ranging between US$33 and $US38; Europe and Central Asia, Middle East and North Africa, and Latin America and the Caribbean seem to be the most expensive, with the cost per surgical DALY averted ranging between US$77 and US$94; and East Asia and the Pacific falls in the middle. This finding indicates that, from the perspective of providing surgical care, a district hospital is an exceptional "buy" in Sub-Saharan Africa and South Asia, both areas with high disease burdens. Coupled with evidence that district hospitals are comparatively underfunded compared with national (tertiary) hospitals (Fiedler, Wight, and Schmidt 1999), a prima facie case exists for increasing support for district hospitals in developing countries. However, those providing such support have to be cognizant of realities on the ground, especially political realities, because they have a significant effect on the direction of change (Blas and Limbambala 2001).

Data on the cost-effectiveness of surgical interventions for specific conditions in developing countries are scarce. One notable exception is for the surgical treatment of cataracts (removal of the opaque lens with or without the insertion of an intraocular implant). Blindness from cataracts is a significant public health problem in many developing countries, and as their populations age, estimates indicate that by 2020 more than 40 million people will be blind or almost blind because of cataracts (Brian and Taylor 2001). Box 67.3 describes a successful program in India.