Cross-sectional analysis tracking workforce density in surgery, anesthesia, and obstetrics as an indicator of progress toward improved global surgical access

Abstract

Introdution: 

Safe surgical care, including anesthesia, obstetrics, and trauma, is an essential component of a functional health system, yet is lacking in much of the world. One indicator of surgical access is the number of specialist surgeons, anesthesiologists, and obstetricians (SAO) per 100,000 population, but global progress reaching threshold SAO density (SAOD) is unknown. This study measured SAOD change/trajectory and highlighted components of workforce expansion.

Methods: 

SAOD in 2019 was captured utilizing publicly available medical licensing data for a convenience sample of 21 countries. Projected 2030 SAOD were estimated by extrapolating annual changes since 2015. Ugandan medical students were surveyed regarding postgraduate plans and SAO training availability. Workforce contribution by nonphysician surgical and anesthetic providers was measured in Sierra Leone.

Results: 

Three low-income countries (LICs), 4 lower middle-income countries (L-MICs), 7 upper middle-income countries (UMICs), and 7 high-income countries (HICs) were included. Overall SAOD increased since 2015. The average 2019 SAOD was 1.16±0.81 (LICs), 3.19±1.92 (L-MICs), 20.98±12.55 (UMICs), and 44.04±12.41 (HICs). The projected 2030 SAOD in LICs and L-MICs remains below 20. In Uganda, 144 specialist SAO training positions and practice preferences predict an inadequate future workforce. In Sierra Leone, nonphysician providers contributed a 6-fold increase in the surgical workforce, though remains inadequate.

Conclusions: 

Despite incremental positive changes since 2015, the current SAOD trajectory is inadequate to realize 2030 access goals. Increased training and retention of specialists and nonphysician providers are necessary to address this critical deficit.

Surgical care, including safe anesthesia, obstetric, and trauma care is a critical component of a functional health system. In 2015, the Lancet Commission on Global Surgery (LCoGS) estimated 5 billion people lack access to safe, timely, and affordable surgical care due to insufficient infrastructure, excessive costs, and a shortage of providers¹. Currently, surgically treatable conditions, including trauma, comprise one-third of the global burden of disease, yet lack of access to treatment results in an estimated 17 million people dying annually and even more suffering significant disability^2,3. In 2015, the World Health Assembly (WHA) passed Resolution WHA68.15, a commitment to strengthen emergency and essential surgical and anesthesia care in low- and middle-income countries (LMICs)⁴. Specifically, it aims to achieve access to safe and timely emergency and essential surgery for 80% of the world’s population by 2030^4,5.

The LCoGS established and measured indicators of access to safe and affordable surgical care in 2015, including specialist workforce density: the number of specialist surgeons, anesthesiologists, and obstetricians (SAO) per 100,000 population^1,6,7. Importantly, SAO density (SAOD) correlates with maternal mortality ratio (the number of maternal deaths per 100,000 live births) and surgical volume^7,8. The LCoGS estimated the SOAD needed to allow for 80% of patients having access to safe and timely surgery by 2030 is at least 20¹. Although surgical workforce density alone cannot guarantee access to quality surgical care, it is a critical component and a key indicator of progress^1,6,8. SAOD is also an easier indicator to track by using annual licensing data for physicians, whereas other indicators are rarely reported⁶.

While identifying the number of current practicing specialists is critical, sustainable change in the workforce depends on training and retaining adequate specialists to provide care. Indeed, the future SAOD depends on adequate numbers of postgraduate training positions in surgery, anesthesia, and obstetrics, support for students to train in SAO specialties and subsequently remain and work in-country, and physician emigration and retirement^1,7–9.

Physician specialists are needed, yet the duration and cost of training specialist physicians and the limited number of students entering SAO specialties in LMICs have led many countries to utilize nonphysician providers, known as task-shifting or task-sharing¹⁰. Task shifting, “the redistribution of responsibilities from highly qualified professionals to those with fewer qualifications,” is a common, though not universal, practice in anesthesia, with anesthesia technicians or nurse anesthetists providing care^1,10. The majority of task-shifting for surgery is to nonspecialist physicians, but some countries also utilize trained surgical technicians to provide operative care¹⁰. The current definition of SOAD includes only specialist physicians, thus despite significant contributions to patient care, task-shifting is not included in SOAD calculations.

Despite the clear consensus goal of 80% access to emergency and essential surgery by 2030, there is a paucity of data on the progress toward this⁶. The World Federation of Societies of Anaesthesiologists (WFSA) workforce survey marked a valuable step for measuring the anesthesia workforce, but without data on progress across the surgical workforce, we cannot determine if and what changes are needed to meet the 2030 goal¹¹. To measure progress, we aimed: (1) to determine the 2019 SAOD for a sample of low-income countries (LICs), lower-middle-income countries (L-MICs), upper-middle-income countries (UMICs), and high-income countries (HICs) and (2) to compare these to 2015 measurements and estimate projected 2030 SAOD based on the current annual rates of change. In addition, we highlight 2 countries to examine the impact of SAO postgraduate training positions and task-shifting on the surgical workforce.

Methods

Study design

A cross-sectional study was conducted to determine the 2019 SAOD for a sample of LICs, L-MICs, UMICs, and HICs dispersed throughout the 6 World Health Organization (WHO) regions with available online medical licensing registers. This study was approved by our hospital’s Institutional Review Board.

Primary outcomes

Primary outcomes included 2019 SAOD numbers per 100,000 population and projected 2030 SAOD in countries below the 20 SAOD benchmark in 2019. Secondary outcomes explored medical students’ postgraduation plans and SAO specialist training positions in Uganda and the impact of task-shifting on the surgical workforce in Sierra Leone.

Data sources and analysis

SAOD data was collected from publicly available, online medical licensing registries by country for 2019 (Appendix I, Supplemental Digital Content 1, https://links.lww.com/IJSGH/A6) from all but 3 countries, whose data was obtained from their national surgical, obstetric, and anesthesia plan (NSOAP) development process (Pakistan and Nigeria) or from recent publication (Colombia)¹². The 2015 SAOD data was obtained from the WHO Global Surgical Workforce Database Collection Tool¹³. In an effort to exclude data aberrant from previous studies but to still allow for temporal variation, 2019 SAOD that differed from the 2015 WHO data by >40% were excluded. World Bank Data for population and income classification using the Gross National Income (GNI) per capita¹⁴. The number of licensed SAO in 2019 was collected from each country’s medical license register. Surgical specialties included cardiothoracic, general, neurosurgery, ophthalmology, orthopedic, otorhinolaryngology/ENT, pediatric, plastic, and urology subspecialties. SAOD was calculated dividing the total number of specialists per country by the 2019 population estimates per 100,000 people^7,8. Each country’s 2019 SAOD was then compared with the SAO benchmarks of 20 or 40 per 100,000 population. The difference between each country’s SAOD in 2015 and 2019 was used to calculate an annual rate of change. Extrapolating this rate allowed estimates of the 2030 SAOD. Descriptive statistics were used to compare country income groups.

The Uganda chapter of InciSioN, an international student and resident global surgery network, conducted a survey to examine medical students’ post graduate plans and SAO training positions in Uganda. Descriptive statistics were conducted to determine the number of and percent filled SAO training positions and students’ postgraduate plans: abroad versus in-country and urban versus rural regions (Appendix II, Supplemental Digital Content 1, https://links.lww.com/IJSGH/A6).

Data on nonspecialist workforce in Sierra Leone was collected for nonphysician anesthesia providers (private communication) and licensed surgical technicians¹⁵. This allowed for comparison of the 2019 SAOD to the combined specialist and nonspecialist surgical workforce.

Results

We sampled 3 LICs, 4 L-MICs, 7 UMICs, and 7 HICs. Sri Lanka, Costa Rica, and Switzerland were excluded because 2019 SAOD were >40% different than 2015 values, indicating likely inaccurate or poor-quality data.

Among LICs sampled, the average SAOD was 1.16 (SD: 0.81) (Table 1). Sierra Leone’s SAO was the smallest at 18, or an SAOD 0.23. SAOD was similar in Rwanda with SAOD 1.69 and in Uganda with SAOD 1.57.

Table 1 - Countries included in analysis, organized by World Bank income bracket, with documented surgeon, anesthesia, obstetric numbers.

Country	Income Bracket	2019 Population (million)	Surgeon Numbers	Anesthesia Numbers	Obstetrics Numbers	SAO (Total)	SAO Density (per 100,000)
Rwanda	LIC	12.63	98	31	84	213	1.69
Sierra Leone	LIC	7.81	8	6	4	18	0.23
Uganda	LIC	44.27	362	72	262	696	1.57
Kenya	L-MIC	52.57	749	202	440	1391	2.65
Nigeria	L-MIC	200.96	2300	254	733	3287	1.64
Pakistan	L-MIC	216.56	—	—	—	—	6.00
Zimbabwe	L-MIC	14.65	169	87	108	364	2.48
Colombia	UMIC	50.34	5388	3200	1647	10,235	20.33
Fiji	UMIC	0.89	29	28	19	76	8.54
Malaysia	UMIC	31.95	2913	1042	1022	4977	15.58
Maldives	UMIC	0.52	51	6	26	83	15.96
Mauritius	UMIC	1.27	249	94	96	439	34.57
Peru	UMIC	32.51	7282	2219	4079	13,580	41.77
South Africa	UMIC	66.63	2912	2526	1298	6736	10.11
Bahamas	HIC	0.39	54	17	39	110	28.21
Ireland	HIC	4.8	891	820	248	1959	40.81
Israel	HIC	8.52	3127	997	1747	5871	68.91
New Zealand	HIC	4.78	964	983	330	2277	47.64
Singapore	HIC	5.8	1524	521	346	2391	41.22
Trinidad and Tobago	HIC	1.4	303	117	135	555	39.64
US Virgin Islands	HIC	0.11	34	5	7	46	41.82

Surgeon, anesthesiologist, and obstetrician densities calculated using World Bank population estimates.

Only aggregate data, not raw numbers, were available for Pakistan.

HICs indicates high-income countries; LICs, low-income countries; L-MICs, lower middle-income countries; SAO, surgeon, anesthesiologist, obstetrician; UMICs, upper middle-income countries.

Among L-MICs sampled, the average SAOD was 3.19 (SD: 1.92) and ranged from 1.64 in Nigeria to 6 in Pakistan. Among UMICs, the average SAOD was 20.98 (SD: 12.55) and ranged from 8.54 in Fiji to 41.77 in Peru. Among HICs, the average SAOD was 44.04 (SD: 12.41) and ranged from 28.21 in the Bahamas to 68.9 in Israel.

For LIC and MICs with 2019 SAOD below the benchmark of 20, the number of SAO specialists needed to meet 2030 goals were estimated using the World Bank’s estimated population growth by 2030 (Table 2). Additional workforce numbers needed to reach SAOD of 20 ranged from an additional 21 SAO providers in the Maldives to 49,309 needed in Nigeria.

Table 2 - Projected workforce needed to meet low-income countries, lower middle-income countries and upper middle-income countries surgeon, anesthesiologist, and obstetrician densities goals by 2030.

Country	Income Bracket	2019 SAO Number	2030 Projected Population (million)	SAO Needed for SAOD of 20/100,000 in 2030 (# Additional Specialists)
Rwanda	LIC	213	16.23	3246 (+3033)
Sierra Leone	LIC	18	9.65	1930 (+1912)
Uganda	LIC	696	59.44	11,888 (+11,192)
Kenya	L-MIC	1391	66.45	13,290 (+11,899)
Nigeria	L-MIC	3287	262.98	52,596 (+49,309)
Pakistan	L-MIC	—	262.96	52,592
Zimbabwe	L-MIC	364	17.6	3520 (+3156)
Fiji	UMIC	76	0.97	194 (+118)
Malaysia	UMIC	4977	36.09	7218 (+2241)
Maldives	UMIC	83	0.52	104 (+21)
South Africa	UMIC	6736	65.96	13,192 (+6456)

HICs indicates high-income countries; LICs, low-income countries; L-MICs, lower middle-income countries; SAO, surgeon, anesthesiologist, obstetrician; SAOD, surgeon, anesthesiologist, obstetrician density; UMICs, upper middle-income countries.

Education: Uganda example

Of the 2393 Ugandan medical students enrolled in 6 medical schools across the country, 141 medical students completed surveys through InciSioN-Uganda (Appendix II, Supplemental Digital Content 1, https://links.lww.com/IJSGH/A6). Of those, 15.6% (n=22) indicated a preference of working in urban areas within Uganda, whereas 7.1% (n=10) showed a willingness to work in rural areas. However, the majority of respondents, 77.3% (n=109), indicated a preference to move abroad to practice.

In 2019, 5 Ugandan training sites (Makerere, Mbarara, Kampala International University, Kozi, and Kabale) had postgraduate positions in SAO specialties, with 144 annual positions. However, the number of annual graduating SAO specialists is lower due to unfilled training positions, often because of lack of funding or new program development. If all positions were filled and trainees remained in Uganda to practice, the surgical workforce would increase by 1584 SAO by 2030. Without accounting for retirements or migration, SAOD would increase from 1.55 in 2019 to 3.81 in 2030. However, current rates of increase would project a density of 2.92 SAO, likely more accurately reflecting local training, retirement, and emigration realities (Figs. 1, 2).

Figure 1: Surgeon, anesthesiologist and obstetrician density for 2015, 2019, and projected 2030 trends in LIC, L-MIC, UMIC and HIC countries. HICs indicates high-income countries; LICs, low-income countries; L-MICs, lower middle-income countries; SAOD, surgeon, anesthesiologist, obstetrician density; UMICs, upper middle-income countries.

Figure 2: Surgeon, anesthesiologist and obstetrician density for 2015, 2019, and projected 2030 trends in LIC and L-MIC countries. LICs indicates low-income countries; L-MICs, lower middle-income countries; SAOD, surgeon, anesthesiologist, obstetrician density.

Task-shifting: Sierra Leone example

Recently suffering from an Ebola outbreak, Sierra Leone has had one of the lowest human development indices since 1982. Unfortunately, this is reflected in the country’s SAO number: 8 surgeons, 6 anesthesiologists, and 4 obstetricians provide care for 7.7 million people, for an SAOD of 0.23. A hundred-fold increase is needed to reach the target of 20 SAOD.

Like other countries, Sierra Leone has utilized task-shifting to address the surgical workforce deficit^1,10. The addition of an estimated 72 nurse anesthetists and 31 surgical technicians trained and working in Sierra Leone has increased the surgical workforce by 6 times^10,11. Nurse anesthesia training is 18 months, and surgical training for Community Health Officers and doctors working in district hospitals is 3 years, including an internship. With 3 classes per year currently enrolled in CapaCare surgical technician training, 34 more trainees are expected to graduate by 2021¹⁵.

Discussion

In the countries sampled, nearly all made positive incremental improvements in SAOD; however, the LICs and L-MICs remain far below the goal of 20 by 2030. The average SAOD in UMICs was relatively stable, though below the SAOD benchmark of 40 for UMICs. HICs have remained stable and are projected to maintain an average SAOD over 40 by 2030. Therefore, while the SAOD trajectory is overall positive, it is clearly inadequate and disproportionately affects LICs and L-MICs, the region’s most critical to achieving the 2030 target of 80% access. This equates to specialist shortages that require increasing the SAO workforce by 10–20 times the current workforce, necessitating a massive scale-up in training and jobs.

Increasing the SAOD in LICs and L-MICs requires a sufficient number of training positions for these specialties. Given the time from entry to medical school to completion of specialist training averages 9–10 years, increased enrollment has a significant lag time. Moreover, encouraging graduates to remain in-country and practice in areas of need is critical to ensure access to timely surgery. In the Ugandan example, the current number of SAO specialty training positions is insufficient to increase the SAOD to 4, let alone to meet the 2030 goal of 20. This is likely the case in many L-MICs¹⁶. Successful interventions to address training shortages in Sub-Saharan Africa include multicountry collegiate training and accreditation programs. The College of Surgeons of East, Central, and South Africa (COSECSA) surgical training, with 12 training sites accredited in Uganda 2019, has significantly increased training opportunities in surgery¹⁷. Similarly, collegiate multicountry programs through the College of Anaesthesiologists of East, Central, and South Africa (CANECSA) and the East, Central, and South African College of Obstetrics and Gynecology (ECSACOG) will soon contribute to Anesthesia and Obstetrics/Gynecology training opportunities¹⁷. Also notable are the responses of medical students interested in SAO specialties, with the majority reporting a preference to practice abroad or in an urban center in-country. Understanding and addressing the reasons behind the practice location preferences is critical to addressing the need to train and retain the SAO workforce in LMICs^18,19. However, emigration realities may not reflect the student’s stated preferences, as surgeons trained within the region actually have a high graduate retention rate, with 88% of surgeons trained in COSECSA programs working in East, Central, and South Africa after graduation²⁰. Overall, these findings indicate there is an urgent need to support and invest in workforce training and retention in LMICs.

In the Sierra Leone example, task-shifting to surgical technicians resulted in a 36% increase in surgical volume and an increase in surgical workforce in hospitals with surgical technicians and nurse anesthetists, while still maintaining high quality care^15,21. However, given population growth, the number of procedures per population has remained stable^14,15. Already utilized in many LMICs, training non-physician providers is faster and less costly than specialist physician training, yet still maintains excellent outcomes for procedures within their scope of practice; therefore, task-shifting should be considered in strategic planning to increase the SAO workforce^22,23. General practice physicians’ contribution to SOA care is also invaluable, with the general duty doctors/medical officers doing the majority of cesarean sections and operations in district hospitals across Sub-Saharan Africa²⁴. Nevertheless, physician specialists are still needed for training, supervision, and performing complex cases and therefore, task-shifting cannot entirely replace the need for specialized physicians^1,10.

A number of other solutions have been proposed to address the shortage of SAO providers. NSOAP development and its implementation should involve all critical stakeholders to strategically allocate resources and to identify urgent needs, including the need for increased workforce¹. Increasing the number and funding of SAO training positions may incentivize trainees to enter the SAO workforce, which will be critical to increase SAOD in LMICs^9,17,25. Supporting new SAO specialists early in their careers through mentorship may also improve retention and reduce the rate of migration to practice abroad²⁵. To increase training opportunities, some have proposed an “encore career” for SAO providers from HICs near retirement, particularly to aid specialty training in lower-resource countries²⁶.

While this study offers an objective assessment of progress towards improving global surgical access, there are notable limitations. The sample was limited to 21countries with publicly available medical licensing registries that document provider specialty. Nevertheless, we included representative countries from all WHO’s regions and different World Bank income classification. For included countries, the licensing data was up to date; however, when there were major differences (>40%) between 2015 data and 2018 data, countries were excluded from the analysis, leading to omission of Switzerland, Costa Rica, and Sri Lanka. This was a minority of the data points included, suggesting that the data accuracy was similar to 2015. Licensing registries generally require yearly or bi-yearly registration, require voluntary specialist designation, and do not always indicate active practice or full-time practice in-country, so listed specialists may be retired, deceased, or practicing abroad most of the year. Defining who is included as a specialist is critical for comparison, yet is not standardized and leads to difficulty utilizing or comparing different sources. For example, medical licensing data from Ireland reveals an SAOD of 40.81, but World Bank data report an SAOD of 71.97 in 2018²⁷. Exploring this discrepancy with the Medical Council of Ireland revealed that the inclusion of all physicians who identify surgery, anesthesia, and obstetrics as an area of practice, including trainees and nonspecialist physicians, results in an SAOD of 79 in 2019. Ireland’s SAOD calculated by the WHO surgical workforce database collection tool, our reference for 2015 comparisons, was 49, closer to the value we identify here^13,27. To facilitate comparison, we followed the strict definition of SAOD as practicing specialist physicians, even though this will not capture all the surgical workforce in any country and particularly misses nonspecialist physicians, nonphysicians, and trainees who provide surgical, anesthetic, and obstetrical services throughout many parts of the world¹. The 2015 WHO data was submitted by country representatives whose methodology for the surgical task force collection may have differed from ours, potentially causing some of the discrepancies with our data and subsequent projections. In addition, the projected 2030 SAOD are based only on 2 data points to calculate an annual rate of change and may misrepresent the actual trajectory. This is likely seen with some of the negative trajectories for countries such as Ireland, Israel, and South Africa and ideally, as countries identify and implement interventions to meet the 2030 SAOD goal, the trajectories may actually be curvilinear. Despite this limitation, we believe a measure of progress is needed, and this study may serve as a template that can be repeated to more accurately determine workforce trajectories. Finally, survey responses should be interpreted cautiously, as they were from a small proportion of Ugandan medical students and may not be widely representative; however, they are likely to include a high proportion of those interested in surgical specialties, given that it was administered to InciSioN-Uganda members—a local branch of the International Student Surgical Network.

Despite these limitations, this study does assess progress in SAOD globally and provides objective data that is actionable for governments and nongovernmental organizations looking to improve surgical access and care. As it stands, the current SAOD are consistent with 2015 numbers and trends are concerning. If major changes are not instituted in the next decade, particularly in LMICs, billions of people will remain unable to access life-saving surgical care. Future analysis should include regular tracking of progress toward achieving SAOD goals and repeat determination of the correlations with maternal mortality ratio, surgical workforce density and surgical volume. Surveying physician trainees more broadly would also help to understand the future contributions to SAOD. As NSOAPs are developed and implemented, comparing SAOD in countries with NSOAPs to those without may help determine their impact on the surgical workforce. Finally, as more surgical volume and outcome data become available, the other LCoGS surgical indicators should be examined to evaluate additional elements of surgical access, with particular attention to trauma care which is poorly assessed by most of the indicators.

Conclusion

This is the first comparison of SAOD around the world before and after the WHA68.15 resolution to achieve access to surgical, obstetric, trauma, and anesthetic care for 80% of the world by 2030. It highlights the need for urgent measures to increase the SAO workforce to address the health care needs of the world’s population. Without major interventions to increase training and retention of SAO specialists in countries around the world, many patients will continue to die and suffer disability due to preventable and treatable surgical illnesses. LICs and L-MICs have extremely low numbers of specialist providers and, while there are increases in SAOD in the countries sampled, the trajectory is clearly inadequate to reach the 2030 goal. Global efforts must include interventions to train specialists and to provide jobs in underserved areas, with consideration to increasing and incentivizing training positions in-country and to utilizing nonspecialist and nonphysician providers to address critical shortages.

Assistance with the study

The authors formed the Tracking and Indicators Working Group of The G4 Alliance, which supported the development of this paper.

Presentation

Preliminary data from this study was presented at the G4 Alliance Semi-Annual Permanent Council Meeting in Manila, Philippines in November 2019. Preliminary data was also presented at the Northwestern Global Health Institute Global Health Day in December 2019.

Ethical approval

The Ann & Robert H. Lurie Children’s Hospital of Chicago IRB approved the protocol of this study and found it exempt.

Sources of funding

No funding was secured for this study.

Author contribution

All authors have approved the final manuscript for submission. M.E.B., J.J., D.V., J.G.-B., A.E., and M.L.: study conception and design. M.E.B., J.J., D.V., J.G.-B., A.E., and M.L.: acquisition of data. M.E.B., J.J., and M.L.: drafting of manuscript. M.E.B., J.J., D.V., J.G.-B., A.E., and M.L.: critical revision.

Conflict of interest disclosures

The authors declare that they have no financial conflict of interest with regard to the content of this report.

Research registration unique identifying number (UIN)

researchregistry5880.

Guarantor

Nothing to disclose.