2006 by the Société Internationale de Chirurgie
Published Online: 8 March 2006
World J Surg (2006) 30: 505–511
DOI: 10.1007/s00268-005-0609-5
Cost/DALY Averted in a Small Hospital in Sierra
Leone: What Is the Relative Contribution of
Different Services?
Richard A. Gosselin, MD, MPH, MSc,1 Amardeep Thind, MD, PhD,2
Andrea Bellardinelli, BA3
1
School of Public Health, University of California, Berkeley, California, USA
Department of Family Medicine, Department of Epidemiology and Biostatistics, Schulich School of Medicine, University
of Western Ontario, 245-100 Collip Circle, London, Ont N6G 4X8, Canada
3
Logistician-Administrator, Emergency Hospital, Goderich, Sierra Leone
2
Abstract
Background: A cost-effective analysis (CEA) can be a useful tool to guide resource allocation
decisions. However, there is a dearth of evidence on the cost/disability-adjusted life year (DALY)
averted by health facilities in the developing world.
Methods: We conducted a study to calculate the costs and the DALYs averted by an entire
hospital in Sierra Leone, using the method suggested by McCord and Chowdhury (Int J Gynaecol
Obstet 2003;81:83–92).
Results: For the 3-month study period, total costs were calculated to be $369,774. Using the
approach of McCord and Chowdhury, we calculated that 11,282 DALYs were averted during the
study period, resulting in a cost/DALY averted of $32.78. This figure compares favorably to other
non-surgical health interventions in developing countries. We found that while surgery accounts
for 63% of total caseload, it contributes to 38% of the total DALYs averted.
Conclusions: Surgical treatment of some common pathologies in developing countries may be
more cost-effective than previously thought, and our results provide evidence for the inclusion of
surgery as part of the basic public health armamentarium in developing countries. However, these
results are highly context-specific, and more research is needed from developing countries to
further refine the methodology and analysis.
C
ost-effectiveness analysis (CEA) is becoming an
increasingly important tool for policymakers in
developing countries. When faced with resource constraints, CEA can be used to facilitate allocation decisions
by policymakers in a manner that maximizes the overall
health of the recipients.1 The World Health Organization
(WHO) has published guidelines for CEA, using the Disability Adjusted Life-Year (DALY) as the unit of measurement of effectiveness.1
Correspondence to: Amardeep Thind, MD, PhD, e-mail: athind2@
uwo.ca
Literature examining cost-effectiveness in developing
countries usually focuses on specific conditions.2–4
Although these studies may be beneficial in deciding the
cost-effectiveness of a disease-focused program (for
example, cataract surgery), policymakers are often faced
with the decision of whether to construct a clinic/hospital or
not. Knowing the cost-effectiveness of such facilities would
be of tremendous benefit in such allocation decisions.
Unfortunately, the literature examining such facilitybased cost-effectiveness is sparse.5,6 The present study
adds to the literature by estimating the cost-effectiveness
506
Gosselin et al.: Cost/DALY Averted in a Sierra Leone Hospital
of a small hospital in Sierra Leone and discusses the
contributions of different services at this hospital, with an
emphasis on surgery.
cataract surgery.3,4 For example, a recent analysis reported that it cost only 57 International dollars per disability-adjusted life year (DALY) averted in the WHO
Southeast Asia region.2
In contrast to these disease specific analyses, only two
studies have attempted to estimate the cost-effectiveness
of an entire surgical facility/ward. McCord et al. evaluated
a small 50-bed hospital in Bangladesh and presented
their results in US$ per DALY averted.6 They found a net
cost of only $10.93 per DALY averted for the entire
hospital activity. Debas et al. in the forthcoming second
edition of the World Bank’s book Disease Control Priorities in Developing Countries estimated the cost-effectiveness of the surgical ward of a typical district hospital in
developing countries. They found that cost per surgical
DALY averted at the district hospital level in sub-Saharan
Africa and South Asia is between $33 and $38; and that it
is between $77 and $94 in Europe and Central Asia, the
Middle East and North Africa, and Latin America and the
Caribbean.5 They concluded that, from the perspective of
providing surgical care, a district hospital is an exceptional ‘‘buy,’’ both in sub-Saharan Africa and in South
Asia, areas with high disease burdens.
This article adds to the literature by adopting the McCord
et al. approach and using actual data from a small hospital
in Sierra Leone to estimate the cost/DALY averted, thus
enabling comparisons to other non-surgical health interventions in resource-poor situations. This research will add
to the evidence available on which policymakers can base
decisions for allocation of scarce resources.
LITERATURE REVIEW
Surgery lies at the end of the spectrum of the curative
medical model. In contrast to simple interventions such as
oral rehydration solution (ORS), the role of surgical interventions in poor and developing countries has traditionally
been considered minimal, because of perceived high costs
and limited human and material capacity availability. It is
thus not surprising that there are only a few studies
examining cost-effectiveness of surgery in developing
countries. For example, the first edition of the landmark
book Disease Control Priorities in Developing Countries
barely mentioned the cost-effectiveness of surgical interventions, save for cataract surgery.7 The forthcoming
second edition, however, has an entire chapter devoted to
cost-effectiveness of surgery in resource-poor environments, which testifies to the growing perception of surgery
as an important part of the public health armamentarium.5
Earlier studies attempted to assess the cost structure of
surgical facilities/hospitals;8–10 none attempted to examine the effectiveness of these facilities. More recent work
has focused on linking these cost analyses with some
dimension of output. For example, Shepard et al. reported on the costs of surgical repair of inguinal hernias at
two types of facilities in Colombia and measured effectiveness in terms of complication rates and patient satisfaction. They found that it cost $39.12 to repair an
inguinal hernia in an intermediate health unit and $148.76
to do so in a hospital, but they did not find any statistically
significant differences in complication rates and patient
satisfaction between these two locations.11
A study from Ghana evaluated the cost-effectiveness of
40 health interventions, including three surgical conditions (severe trauma, appendicitis, and hernia).12 The
authors reported their results in terms of US$ for costs
and Life Years Saved (LYS) for effectiveness, and found
that cost-effectiveness of appendectomy was $36/LYS,
whereas it was $74/LYS for hernias and $233/LYS for
severe trauma. In contrast, medical treatment for diarrhea
was $74/LYS and treatment of malaria was $84/LYS.
From a surgical perspective, it is important to note that
these authors did not define whether the surgical treatment of hernias was emergent or elective; moreover, they
did not define what constituted ‘‘severe’’ trauma.12
The only surgical intervention that is well proven and
accepted to be cost-effective in developing countries is
METHODS
The Setting
Sierra Leone is a small country in western Africa, with
an estimated population of 6 million people. Although a
devastating civil war officially ended in 1999, the country is
still struggling to re-establish its institutions and the rule of
law. It has consistently ranked at the bottom on the United
Nations’ Human Development Index (HDI), which is
based, among other factors, on a country’s life expectancy, per capita gross domestic product, and maternal
and under-five mortalities. The national health care system is practically non-existent; government hospitals and
other health care facilities (what is left of them) are
essentially non-functional because of severe lack of human and material resources. There are a few private
clinics or hospitals in the capital, which cater to expatriates
or the handful of local nationals who can afford them; by
Gosselin et al.: Cost/DALY Averted in a Sierra Leone Hospital
507
and large it is the non-governmental organizations
(NGOs) that provide health care to the population.
Emergency Hospital is an Italian NGO dedicated to
helping civilian victims of wars and conflicts. It built and
has operated a surgical hospital in the capital city of
Freetown since 2001. This hospital provides, at no cost to
patients, general and orthopedic surgical care, and has
recently expanded its services to provide pediatric outpatient and inpatient care as well. It is the only functioning
surgical hospital in the country, and is a referral center to
which NGOs working in remote areas routinely transfer
surgical cases. Of note is the fact that the hospital does
not provide any obstetric services.
The hospital has 90 beds, 60 of which are in surgical
wards (30 pediatric and 30 adult beds), 16 in the pediatric
inpatient medical ward, and the rest divided between the
intensive care unit (ICU) and for patients with spinal cord
injuries. The hospital usually functions at 90%–100% of
capacity, and extra beds are added when this capacity is
exceeded. The hospital has two operating rooms, a laboratory with a blood bank, an x-ray department, a pharmacy,
a physiotherapy department, and support staff for administration, kitchen, laundry, maintenance, and transportation. The surgical outpatient department (OPD), which
includes an area for dressing changes, sees an average of
60 patients and admits an average of 8 patients a day
(including elective admissions for the following day). The
pediatric OPD sees an average of 30 patients a day, of
which an average of three are admitted.
Staffing is fairly consistent throughout the year. For the
study period (July–September 2004), there were 9 expatriate and 177 full-time national staff. The expatriate staff
included a program coordinator, a medical coordinator, a
general surgeon, an orthopedic surgeon, an administrator,
a physiotherapist, and three nurses. The national staff
comprised of 4 physicians (2 pediatricians and 2 surgeonsin-training), 3 nurse-anesthetists, 80 care-related personnel (nurses, therapists, etc.), and 90 non-care-related
personnel (administration, security, cleaners, drivers, etc).
This study was carried out in October–November 2004
by abstracting all hospital ward and departmental records
for the previous three months (July–September 2004).
Based on the authors’ experience, this 3-month period
was considered to be representative of the yearly activity
of the hospital, both in terms of the cost and output
estimation.
category. Fixed costs represent the depreciated monthly
cost of land purchase and hospital construction, and initial equipment (furniture, medical equipment, vehicles,
generators, etc.) required for starting hospital services.
The building construction cost was straight line depreciated to zero over 30 years; a 10-year period was used to
depreciate equipment costs.
Local operating costs included the salaries and benefits
for expatriate and local staff, cost of locally purchased
equipment, consumables and drugs, maintenance costs
for the building and equipment, transportation, fuel (for
vehicles and hospital generators), utilities, and miscellaneous costs (cleaning materials, bank charges, office
equipment, etc.). A non-local operating cost that the
hospital incurs is the cost of a shipping container sent
twice a year from Italy, containing equipment, supplies,
and consumables. The prorated monthly cost of these
deliveries was included in the operating cost of the hospital. We used the exchange rate prevailing in September
2004 to convert leones to USD (2800 leones = 1 US$).
Cost Estimation
We estimated the fixed and operating costs for the
entire hospital and calculated monthly amounts for each
DALY Estimation
We calculated the DALY for each patient seen and
treated in the hospital during the 3-month study period.
Data were abstracted from admission logs, ward and
department logs, and charts of all patients seen and
treated at the hospital, either as inpatients or as outpatients. We used the approach of McCord et al. in calculating DALYs, but with slightly simplified estimates of risk
of death or disability, and effectiveness of treatment.6
Severity of disease was given a weight of 1.0 if the
disease was considered to be fatal >95% of the time
without treatment. A weight of 0.7 was given for those
conditions fatal >50% but <95% of the time, 0.3 for those
fatal <50% but >5% of the time, and 0.0 for those <5% of
the time. Similarly, the effectiveness of treatment was
given a weight of 1.0 if treatment had a >95% chance of a
permanent cure of the given condition, 0.7 if that chance
was <95% but >50%, 0.3 if it was estimated to be <50%
but >5%, and 0.0 if <5%. Years of life lost (YLLs) were
calculated using the discounted numbers provided in the
Global Burden of Disease (GBD) study.13 For the years
lived with disability (YLDs), weight values from the GBD
study were used when available (for example, fractures,
burns, amputations, diarrheal diseases, and anemia/malaria); when these were unavailable in the published
literature, the authors used their own estimates. Appendix
1 depicts the actual calculations for specific cases.
DALYs averted among ICU patients were attributed to
the inpatient ward where the patients were finally
508
Gosselin et al.: Cost/DALY Averted in a Sierra Leone Hospital
Table 1.
Monthly operating costs
Category
Fixed costs
Land purchase and hospital construction
Equipment costs
Operating costs–local
Expatriate staff–salaries and benefits
Local staff–salaries and benefits
Medical equipment, consumables, and drugs
Maintenance of building and equipment
Transportation
Fuel
Utilities
Miscellaneous
Operating costs–non-local
Container from Italy
Total
US$
1597
6731
38,641
28,741
3899
4921
2315
10,355
2078
10,651
13,329
123,258
transferred. No DALYs were attributed to the few ICU
patients discharged directly home, as these were cases
admitted to the ICU (as the hospital was full) for overnight
observation, for a condition that ultimately proved to be
benign. The same reasoning was applied to the physiotherapy department, which did not contribute to DALYs
averted because the outpatient department sees patients
that are discharged from the wards or referred from
OPDs.
The cost per DALY averted was calculated by summing
the fixed and operating costs of the hospital for the
3-month study period and dividing it by the sum of all
DALYs averted during the same 3-month period.
RESULTS
The cost for land acquisition and hospital construction
was $575,000, and an additional $807,760 was spent on
equipping the hospital with furniture, medical equipment,
vehicles, and generators to make it fully operational. The
depreciated monthly costs for these factors are $1597
and $6731, respectively. Table 1 lists additional details of
the monthly costs for running the hospital. The three
largest expense categories are expatriate staff, national
staff, and the cost of the shipping container sent from Italy
and its contents, equipment, supplies, and consumables.
During the 3-month study period, the hospital treated
8598 patients, 61% of whom were seen in the surgical
outpatient department (OPD), and 29% in the pediatric
OPD. Inpatient cases accounted for only 10% of the total
caseload (Fig. 1). A total of 11,282 DALYs were averted
during the 3-month study period. The majority of the
DALYs averted (62%) were contributed by the pediatric
inpatient (33%) and the pediatric OPD (29%), with the
surgical service accounting for slightly more than a third
(38%) of the total.
Details of the cases seen and DALYs averted by each
service are shown in Table 2. The surgical OPD saw the
bulk of the cases in the hospital (n = 5252; 61% of total
hospital caseload) but contributed only 998 DALYs (9% of
total DALYs averted). Most of the patients seen here
needed dressings, suture removal, and medical review,
but the DALYs were contributed by fracture reductions
and hernia reductions. The general surgery inpatient
service saw 147 patients (2% of total), but contributed
15% of the DALYs averted, mostly from surgery for acute
abdomen and hernias. The orthopedic inpatient service
saw 5% of the total caseload and contributed 14% of
DALYs averted, primarily from fracture reductions,
amputations, burns, and wound debridements. The
pediatric service, which accounted for less than a third of
the total hospital caseload (32%), accounted for the
majority of DALYs averted (62%), primarily because of
treatment of young children for malaria, diarrhea/dehydration, and acute respiratory infection.
Dividing the total costs for the 3 months (3 · $123,358) by
the total DALYs averted during these 3 months (11,282),
gives the cost per DALY averted, which is $32.78.
DISCUSSION
Quantifying effectiveness is not as straightforward as
measuring costs, and certainly it was more controversial.
We agree with McCord et al. that, although imperfect, the
DALY is the best composite measure available.6 Since its
original description by Murray and colleagues, the use of
the DALY has been widely promoted by the World Bank
and the World Health Organization.13,14 It is the unit of
measurement used in two of the most influential publications in the field, the Global Burden of Disease study
and the Disease Control Priorities in Developing Countries book.7,14 It is beyond the scope of this study to debate the merits and drawbacks of the DALY, as they are
amply documented elsewhere.15–17
Our cost/DALY averted ($32.78) is higher than that
reported by McCord et al. from Bangladesh ($10.83).
However, even at these rates, the hospital compares
favorably with other non-surgical interventions such as
Vitamin A distribution ($9), measles immunization ($15),
or oral rehydration solution (ORS) treatment for diarrhea
($35), as calculated by Jamison.7 Our study strongly
Gosselin et al.: Cost/DALY Averted in a Sierra Leone Hospital
509
Table 2.
Cases seen and DALYs averted by service (July–Sep 2004)
Service
Surgical outpatient department
Dressings and suture removal
Fracture reduction and casting
Medical review
Wound suturing
Hernia reduction
Others/unknown
Subtotal
Surgical inpatient ward (General Surgery)
Acute abdomena
Herniab
Mastectomy
Hysterectomy
Thyroidectomy
Othersc
Subtotal
Surgical inpatient ward (Orthopedic Surgery)
Woundsd
Burnse
Contracture release
Arthodesis/osteotomy/sequestrectomy
Fracturesf
Amputationsg
Othersh
Subtotal
Pediatric OPD
Malaria
Diarrhea/dehydration
Acute respiratory infection
Others
Subtotal
Pediatric inpatient ward
Malaria
Diarrhea/dehydration
Others
Subtotal
Grand total
Cases
3224
338
1157
52
39
442
5252
DALYs averted
785
213
998
998
38
26
14
10
5
54
147
1019
450
24
23
10
174
1700
71
46
27
65
113
15
65
402
216
343
172
531
227
96
1585
1261
609
375
253
2498
1633
789
486
330
3238
222
57
20
299
8598
2247
1361
153
3761
11,282
a
Includes appendicitis, peritonitis, bowel obstruction, trauma, etc.
Includes inguinal and/or strangulated hernias.
c
Includes lipoma, keloid, perineal gangrene, thoracotomy, gastrostomy, pyloroplasty, imperforate anus, malaria, cellulitis, etc.
d
Includes debridement, delayed primary closure, skin grafting, abscess incision and drainage, etc.
e
Includes debridement, skin grafting, etc.
f
Includes conservative treatment, open reduction and internal fixation, external fixation, etc.
g
Includes above- and below-knee amputations.
h
Includes finger and toe amputations, below-elbow amputations, tendon repair, foreign body removal, osteomyelitis, head injury,
spinal injury, soft tissue wounds, polio contracture, bone tumors, etc.
OPD: outpatient department.
b
supports the notion put forward by McCord et al. that the
commonly held dogma in public health circles, that surgery is not cost-effective, particularly in developing
countries, needed to be revisited.
There are a few probable hypotheses why our cost/
DALY averted was higher than that reported by McCord
et al. As reported in Methods, we slightly modified the
model used by McCord et al., specifically by simplifying
the weights. We found that most of the DALYs averted
were contributed by the pediatric OPD and inpatient ward,
as opposed to the surgical service (which contributed 60%
of the DALYs averted in the study of McCord et al.). This
510
Gosselin et al.: Cost/DALY Averted in a Sierra Leone Hospital
Figure 1. Distribution of cases
and DALYS averted.
can be explained in part because our hospital does not
provide obstetric care, whereas in their Bangladesh hospital, obstetrics and gynecology contributed greatly to the
caseload. A successful surgical intervention there often
affected the survival of two patients: a young mother, and
a newborn. Life-saving procedures, particularly in the
young, averted many more DALYs than interventions that
prevented or improved disability. In essence, the case mix
of the two hospitals is quite different.
In our study, although the surgical service provides the
bulk of the caseload, the maximal DALYs are averted in
the pediatric service. This can be explained by the
younger age of the patients, the overall higher severity of
disease, and the higher rate of success with treatment
compared to the surgical service.
It is important to note that we have been highly conservative in calculating the DALYs averted. For example,
on the surgical side, we have not attributed any DALYs to
surgery for simple repair (prophylactic) of inguinal hernias,
head injuries, spinal injuries, sequestrectomies, arthrodesis, and osteotomies. For the pediatric OPD, we calculated DALYS averted using the following assumptions:
<5% would die without treatment, and treatment is
successful between 50% and 95% of the time. It is quite
likely that averting 3238 DALYs in 2498 pediatric patients
is a significant underestimation. Furthermore, and probably as important, all patients had only one diagnosis recorded in the log books. No DALYs averted were
attributed to co-morbidities. Some patients logged as
‘‘head injury’’ also had fractures that were treated. Some
children who underwent sequestrectomies were also
treated for malaria or anemia. All these factors combine to
form a very conservative estimate of DALYs averted in our
study.
In conclusion, despite differences in opinion regarding
the moral and ethical considerations of using DALYs for
such analyses, and the assumptions underlying the calculations, we feel that cost/DALY averted can be a useful
(but not the only) tool in assisting policymakers in making
resource allocation decisions. Our results indicate that the
cost/DALY averted by a hospital in Sierra Leone compares
favorably with the costs of other interventions documented
in the literature. Given that this facility is providing primarily
surgical and pediatric services, our results suggest that
surgery should be included in the basic public health
armamentarium in developing countries.
APPENDIX 1: EXAMPLE OF DALY
ESTIMATION
A 15-year-old male with peritonitis who undergoes
successful surgery (>95% mortality without treatment,
>95% chance of permanent cure) 36.80 (YLL value for
males aged 15) · 1.0 · 1.0 = 36.80 DALYs averted by
surgery.
A 5-year-old female successfully treated for very
severe anemia/malaria (>95% mortality without treatment, <50% chance of permanent cure of the condition)
36.59 (YLL value for females aged 5) · 1.0 · 0.3 = 11
DALYs averted by treatment.
Gosselin et al.: Cost/DALY Averted in a Sierra Leone Hospital
511
A 20-year-old male who survives a road traffic accident
with a unilateral below-knee amputation (<95%
mortality without treatment, >95% chance of permanent
disability, with weight of 0.281) 35.02 · 0.7 · 1.0 ·
0.281 = 6.9 DALYs averted by surgery.
5. Jamison D, Evans D, Alleyne G, et al., editors. Disease
Control Priorities in Developing Countries. 2nd ed. Washington, DC, World Bank, Forthcoming 2006.
6. McCord C, Chowdhury Q. A cost effective small hospital in
Bangladesh: what it can mean for emergency obstetric
care. Int J Gynaecol Obstet 2003;81:83–92.
7. Jamison DT, International Bank for Reconstruction and
Development. Disease Control Priorities in Developing
Countries. New York, Oxford University Press for the World
Bank, 1993.
8. Mills AJ, Kapalamula J, Chisimbi S. The cost of the district
hospital: a case study in Malawi. Bull WHO 1993;71:329–
339.
9. Lewis MA, La Forgia GM, Sulvetta MB. Measuring public
hospital costs: empirical evidence from the Dominican
Republic. Soc Sci Med 1996;43:221–234.
10. Malalasekera AP, Ariyaratne MH, Fernando R, et al. Cost
accounting in a surgical unit in a teaching hospital—a pilot
study. Ceylon Med J 2003;48:71–74.
11. Shepard D, Walsh J, Munar W, et al. Cost-effectiveness of
ambulatory surgery in Cali, Colombia. Health Policy Plan
1993;8:136–142.
12. Jha P, Bangoura O, Ranson K. The cost-effectiveness of
forty health interventions in Guinea. Health Policy Plan
1998;13:249–262.
13. Murray CJ. Quantifying the burden of disease: the technical
basis for disability-adjusted life years. Bull WHO 1994;72:
429–445.
14. Murray CJL, Lopez AD, Harvard School of Public Health,
World Health Organization., World Bank. The global burden
of disease: a comprehensive assessment of mortality and
disability from diseases, injuries, and risk factors in 1990
and projected to 2020. Cambridge, MA: Harvard School of
Public Health for the World Health Organization and the
World Bank; Distributed by Harvard University Press, 1996.
15. Arnesen T, Nord E. The value of DALY life: problems with
ethics and validity of disability adjusted life years. BMJ
1999;319:1423–1425.
16. Bastian H. A consumer trip into the world of the DALY
calculations: an Alice-in-Wonderland experience. Reprod
Health Matters 2000;8:113–116.
17. Reidpath DD, Allotey PA, Kouame A, et al. Measuring
health in a vacuum: examining the disability weight of the
DALY. Health Policy Plan 2003;18:351–356.
The above calculations were undertaken to estimate the
DALYs averted for all patients seen during the 3-month
period. Calculations were done individually for all inpatients. Because patients keep their own outpatient charts,
the available data from OPD are less explicit, but age, sex,
diagnosis, and treatment were retrievable from the OPD
log books. Manual reduction of sedated patients with
incarcerated inguinal hernias, for example, was successful 39 times. These patients had a mean age of 37 years,
and we estimate that >50% would eventually have died
without any treatment and that the treatment is successful
(i.e., definitive) in <50% of the cases. DALYs were calculated as follows: 39 · 26 (YLL value for males aged 37)
· 0.7 · 0.3 = 213 DALYs averted. The same principles
were applied for fractures treated in OPD: 338 patients, of
mean age 24 years, with the assumption that 10% of these
would have a greater than 50% chance of some form of
permanent disability. This averted 785 DALYs, which for
338 patients of mean age 24 is probably an overly conservative estimate.
REFERENCES
1. Adam T, Baltussen R, Tan Torres T, et al. (2003) Making
Choices in Health: WHO Guide to Cost Effectiveness
Analysis. Geneva, World Health Organization.
2. Baltussen R, Sylla M, Mariotti SP. Cost-effectiveness
analysis of cataract surgery: a global and regional analysis.
Bull WHO 2004;82:338–345.
3. Marseille E. Cost-effectiveness of cataract surgery in a
public health eye care programme in Nepal. Bull WHO
1996;74:319–324.
4. Singh AJ, Garner P, Floyd K. Cost-effectiveness of publicfunded options for cataract surgery in Mysore, India. Lancet
2000;355:180–184.