Volume of Neurosurgical Disease

We began by estimating the incidence of 10 conditions encountered by neurosurgeons that form the foundation of essential neurosurgical care: brain and spinal tumors, hydrocephalus, traumatic brain injury (TBI), traumatic spinal injury (TSI), neural tube defects, stroke, CNS vascular anomalies, CNS infections, and epilepsy. Conditions were identified by ranking commonly treated neurosurgical conditions described by an international panel of neurosurgical providers (see Proportion Requiring Neurosurgery below). Then, those conditions in which treatment neglect would directly result in severe disability or death were designated by this report as essential neurosurgical conditions. Notably excluded from the final list are degenerative spine disease and osteoporotic vertebral fractures.¹⁵ The proportion and volume of patients succumbing to severe disability or death if these conditions are not treated surgically is difficult to determine, particularly in resource-limited settings. Due to the recognition of limited capacity, this classification scheme of essential neurosurgical care also excluded important conditions such as pain disorders, spasticity, movement disorders, and other conditions for which neurosurgical treatment might offset disability and lost productivity. Competency with the treatment of the 10 conditions above, however, would equip a neurosurgeon anywhere with the ability to treat the vast majority of patients with neurosurgical disease, and thereby most efficiently avert severe disability or premature death.

Reliable epidemiological figures were available for two diseases (neural tube defect and stroke; http://ghdx.healthdata.org/gbd-results-tool),¹⁰ thus our attention was focused on the remaining conditions. National surgical registries were available for brain and spinal tumors allowing region-specific estimation of disease incidence.³ The Institute for Health Metrics and Evaluation has published epidemiological figures for road traffic injuries in each country, which were used to extrapolate the incidence of TBI. The proportion of traffic collisions resulting in TBI was calculated via a meta-analysis of published traffic injury data. Similarly, the proportion of TBI from which traffic collisions is the mechanism of action was calculated via a meta-analysis. Applying these ratios to the incidence of road traffic injuries delivered an estimation of total TBI from all causes.⁶

For the remaining conditions (hydrocephalus, TSI, CNS vascular anomalies, CNS infections, and epilepsy), an exhaustive systematic review was conducted to identify studies reporting epidemiological data from large, population-based cohorts. After extracting case numbers and sample size, a meta-analysis was performed to estimate disease incidence figures for each disease by WHO region and by World Bank income group. Finally, incidence figures were multiplied by population estimates from the WHO to generate estimates for the annual volume of disease globally. Detailed methodology and results are described in detail elsewhere.^{3,7–9,11,16,19}

Proportion Requiring Neurosurgery

Identifying the incidence of neurosurgical disease requires an estimate of the proportion of neurological injury or illness warranting surgical intervention. Some neurosurgical diseases are best managed operatively, while others require nonoperative neurosurgical management. Some neurological conditions do not require neurosurgical expertise for optimal treatment. While surgical logs from ideal health care systems offer reliable estimates for surgical proportion (i.e., the fraction of neurological conditions requiring surgery), their relevance may not be generalizable to health care systems in differing settings. Furthermore, we learn very little about the proportion of disease benefiting from nonoperative neurosurgical expertise or consultation. Thus, following the precedent established by the Lancet Commission on Global Surgery, we sought expert opinion from a broad, diverse group of surgical providers.

Via electronic survey, nearly 200 neurosurgeons from more than 50 countries were asked to estimate the proportion of specific neurological disorders that, in an ideal world, would require either neurosurgical operation or consultation. To gauge degree of certainty, surgeons indicated how confident they were in their estimation for each disease. Finally, we sought to identify the number of cases that can be reasonably performed each year by the average practicing neurosurgeon. Surgeons indicated their annual case volume and types of cases most frequently performed, then indicated how they perceived their workload at that clinical volume. The full questionnaire and analytical methodology are described in detail elsewhere.⁵

Neurosurgical Workforce

Determination of the current neurosurgical workforce was quantified via a large, multinational global mapping project facilitated by the World Federation of Neurosurgical Societies (WFNS), the WHO, and the Program in Global Surgery and Social Change (PGSSC) at Harvard Medical School. Briefly, Ministries of Health, officers of the 130 societies registered with the WFNS, and the members of the WHO Global Initiative for Emergency and Essential Care were contacted via email for information. For some countries an internet search was conducted to identify the contact information for active neurosurgeons willing and able to complete an online questionnaire on capacity. Surgeons registered with the open access Global Neurosurgery forum (https://globalneurosurgery.org/members) were also included. The data collection instrument captured surgeon numbers and geographic locations, as well as the presence of basic equipment germane to neurosurgical care: high-speed drills, bipolar electrocautery, imaging infrastructure, etc. For the purposes of the current summary, we aggregated the number of neurosurgeons by country and by WHO region to facilitate quantification of the current surgical capacity. A comprehensive description of the methodology and results of the PGSSC workforce mapping initiative is described elsewhere (unpublished data).¹⁴

Estimating the Deficit in Neurosurgical Care

The unmet need, or deficit in neurosurgical care, represents the difference between the total estimated number of essential neurosurgical conditions presenting for treatment and the current number of neurosurgical cases that can be performed given the existing workforce. The number of essential neurosurgical conditions was calculated by taking the product of neurosurgical volume (from each meta-analysis by disease entity^{3,7–9,11,16,19}) and surgical proportion (from the provider survey⁵). The number of neurosurgical cases that can be currently performed, on the other hand, was obtained by taking the product of the number of neurosurgeons (from the Workforce estimate [unpublished data]) and the median number of cases performed (from the provider survey estimate⁵). The deficit was then calculated on a regional and income-group basis to provide a more granular description of need and volume, and to highlight disparity in neurosurgical care across regions. Then, a country-specific ratio of neurosurgeons to essential neurosurgical cases (both surgical and consultative) was estimated. The total number of cases within a region was allocated to each country according to the population proportion of a given country within its WHO region and World Bank income group. For the worldwide estimate, all countries with available workforce data were included, irrespective of WHO or World Bank classification. Incorporating the country-specific neurosurgeon figure from the workforce survey, maps were then created to depict this ratio using Tableau software.

Neurosurgical Cases by Region and Income Group

An estimated 13.8 million essential neurosurgical cases develop each year, of which more than 80% arise in low- and middle-income countries (Table 1). A detailed breakdown of neurosurgical volume by disease can be found in the Appendix (Tables S1–S23). Corresponding to the regional population proportions, an expected 3.5 million and 3.7 million new cases are expected in Southeast Asia and the Western Pacific, respectively. Africa is expected to endure nearly 2 million neurosurgical cases, in contrast to about 665,000 cases in the US and Canada. Europe, the Eastern Mediterranean, and Latin America each will encounter between 1.1 and 1.8 million new neurosurgical cases annually (Table 1). Among the conditions we have defined as requiring essential neurosurgical care, surgery for TBI (burr holes, craniotomy/craniectomy, etc.) accounts for 45%, cerebrovascular accident for 20%, hydrocephalus for 7%, and brain tumors for 5%. Vascular anomalies (2.2%), neural tube defects (0.3%), and spinal tumors (0.1%) occupy a relatively modest proportion of the global neurosurgical need.

The number of cases requiring neurosurgical consultation—but not necessarily surgical intervention—is far greater, approaching 22.6 million annually (Table 2). Again, Southeast Asia and the Western Pacific bear the greatest consultative demand with 5.8 million and 6.2 million cases each year, respectively. Low and middle-income countries (LMIC) endure the greatest burden at 17.6 million cases, compared with 4.3 million cases in high-income countries (HIC).

Neurosurgeons by Region and Income Group

We identified 49,940 practicing neurosurgeons worldwide (unpublished data). Ninety-eight percent (98.9%) of neurosurgeons were identified via the WFNS-sponsored electronic survey, while 560 neurosurgeons (1.1%) from 34 countries (14.9%) in which no data were available were imputed via statistical modeling. The greatest population of neurosurgeons resides in the Western Pacific region, with China and Japan alone accounting for more than 18,000 neurosurgeons. Europe and the US/Canada have the next largest populations of neurosurgeons, with 10,719 and 5296 neurosurgeons, respectively. In the African region, with a population of 990 million, 488 neurosurgeons were identified (Table 3). Forty-four percent (44%) of neurosurgeons worldwide reside in high-income countries.

Previously, we have identified the mean neurosurgical volume reasonably managed by a neurosurgeon in a given country as 223 cases per year.⁵ This assumes a surgeon’s perceived workload is maintained at a level of 75 out of 100 (where 0 is “not at all busy” and 100 is “extremely busy, overworked”). The greatest total neurosurgical capacity exists in the Western Pacific region, where more than 22,000 neurosurgeons are capable of performing nearly 5 million essential neurosurgical cases annually. In contrast, the nearly 500 neurosurgeons in Africa can be expected to perform approximately 110,000 neurosurgical operations. Worldwide, and ignoring geographic boundaries and uneven case distribution, the overall surgical capacity is approximately 11 million essential neurosurgical cases per year (Table 3).

The Deficit in Neurosurgical Care

The current deficit in neurosurgical care is approximately 5.2 million (Table 3). This deficit derives from the difference between the number of essential cases encountered annually and the essential surgical capacity, where essential surgical capacity excludes nonessential cases performed by neurosurgeons, particularly in HIC. To meet this demand, an estimated additional 22,626 neurosurgeons would need to be trained. The greatest deficit resides in Southeast Asia where nearly 2.5 million essential cases go unmet. The tremendous shortage of neurosurgeons in India—where 3500 neurosurgeons are responsible for a population of more than 1.2 billion—is largely responsible for this health care discrepancy. For the entire region, 2.5-times the existing number of surgeons are needed to provide appropriate coverage. In Africa, where more than 1.8 million cases are not being addressed by capable neurosurgeons, the surgical workforce must increase by more than 1700% to meet the regional demand. The vast majority of the population in Africa has no access to a neurosurgeon, independent of other geographic, social, or financial barriers that may exist.¹⁴ In terms of essential care, three WHO regions—the region of the Americas restricted to the US and Canada only, the European region, and the Western Pacific region—do not demonstrate a workforce deficit, although the need for additional neurosurgeons in these regions to meet neurological conditions not accounted for in our analysis remains a salient consideration.

Ongoing Efforts to Increase Neurosurgical Capacity

A number of efforts to increase neurosurgical capacity in LMICs have already begun to make a difference. Enabling nonneurosurgeons to perform neurosurgical interventions (task-shifting) that are technically straightforward and have a high benefit to risk ratio (such as emergency evacuation of an epidural hematoma) is rational. Likewise, teaching nonneurosurgeons to place shunts for hydrocephalus has been advocated by some, but infant hydrocephalus is not immediately life-threatening; although placing the initial shunt is not difficult, the significant complication rate requires timely neurosurgical expertise for subsequent management. The ultimate goal should be to increase neurosurgical capacity by increasing the number of competent and accessible neurosurgeons in underserved regions. The most sustainable way to achieve this will be for adequately trained national neurosurgeons to begin training programs in their home countries. This may initially require the importation of expertise, but the need for this would diminish over time if an investment is made in training individuals from these countries.

The global neurosurgery community should prioritize the provision of training to individuals from countries that do not have adequate neurosurgery residency programs. Historically, people who leave home to train in a high-income country often fail to return. But in the past decade, sub-Saharan Africa has benefitted from an influx of neurosurgeons because the few established training programs on the continent have embraced the opportunity to incorporate trainees from the less developed countries into their residency programs. Programs in Morocco, Senegal, Zimbabwe, and South Africa have trained approximately 60 new neurosurgeons from other African countries thus far. Also, a new East African training program has recently been established in cooperation with The College of Surgeons of East Central and Southern Africa, with training sites in Tanzania, Uganda, and Kenya. All these programs provide more appropriately contextualized training in regard to pathology and resource availability, and are more likely to see their trainees return home to build neurosurgical capacity in their own countries. In addition, longitudinal partnerships between centers in LMICs and academic neurosurgery departments in HICs (Duke University, University of Alabama, University of Miami, The University of Toronto, The Weill Cornell Medical College, The Barrow Neurological Institute, and others) are fostering continued training, mentoring, and collaborative research in sub-Saharan Africa, Southeast Asia, and Latin America. Other programs, such as CURE Hydrocephalus and Spina Bifida, provide subspecialty training fellowships to neurosurgeons from across the developing world. Furthermore, organized neurosurgery (such as the Foundation for International Education in Neurosurgery, the WFNS, the Asian Australasian Society of Neurological Surgeons, the American Association of Neurological Surgeons, the European Society of Pediatric Neurosurgery, and the International Society of Pediatric Neurosurgery) has long been engaged in international neurosurgery education. The past 10 years have seen a noticeable escalation in awareness of and commitment to global neurosurgery that should, if we stay on mission, gradually improve access to neurosurgical care for people everywhere.

The Way Forward

Strengths and Limitations of the Study

The estimates and conclusions outlined in this summary must be considered in the context of several limitations. First, most disease volume figures derive from large systematic reviews and meta-analyses, as reliable, population-based estimates for each country or region are nonexistent. Similarly, the surgical workforce figures relied upon accurate neurosurgical association reporting, and at times questionnaire responses and statistical modeling, all of which carry bias and intrinsic inconsistencies. Throughout the tables and text, we assume that neurosurgeons are optimally distributed geographically in relation to the at-risk population. We also assume that each trained neurosurgeon is capable of safely treating each condition encompassed within the designation of essential neurosurgical care. While these assumptions are generally safe for epidemiological modeling at the region- or country-level, they would not permit accurate allocation of resources at the city or hospital level. In relation to surgical capacity, we have focused on the surgeon only—not the wealth of infrastructure, equipment, and personnel required alongside the surgeon to facilitate safe and timely provision of neurosurgical care. We acknowledge the tremendous importance of these resources, and we are actively conducting a worldwide survey to better understand their accessibility to current and future surgeons.¹⁴

In an effort to prioritize the most life-threatening neurosurgical conditions worldwide, we have deliberately excluded many important conditions commonly addressed by neurosurgeons, most notably degenerative spine disease. This relative limitation warrants specific elaboration. In the US, with more than 5000 neurosurgeons, we estimate the case capacity is nearly 1.2 million cases/year. Meanwhile, the volume of essential neurosurgical cases is estimated at only 576,000, suggesting a surplus of neurosurgeons exists in the US. We would caution the reader against this conclusion. The majority of neurosurgical cases performed in the US are for conditions falling outside the designation used in this report of essential neurosurgical care. Indeed, entities like degenerative spine disease, chronic pain, movement disorders, and many other conditions are well known to cause disability and suffering if left untreated. Thus, a shortage of neurosurgeons may exist even in the most developed of countries, depending upon the level and type of neurosurgical care demanded by a given population. The scope and purpose of this paper is to offer a description of the workforce capacity and deficit that exists for essential neurosurgical care, a designation that itself aims to prioritize those patients who are most likely to benefit from lifesaving and disability-averting neurosurgical care.

Finally, this report ignores the concept of task sharing and the number of procedures potentially performed by nonneurosurgeon providers. Training general surgeons in common neurosurgical procedures such as ventriculoperitoneal shunt insertion and burr hole craniotomy is a practical strategy that might more quickly help curb the overwhelming neurosurgical burden. Reasonable neurosurgical experts will disagree on the ethics and quality-related implications of this frequently used strategy in the realm of global health. The potential capacity-related consequences of this strategy, however, are absent from the figures provided above.

While inherent limitations exist for an epidemiological survey and capacity estimation of this scope, careful and deliberate measures were taken to ensure data validity and provide methodological transparency throughout. Rather than relying on surgical log data from hospitals and health systems where the surgical care is known to be inadequate, we approached surgical deficit figures first from the disease component of the equation. Though data modeling was necessary in some cases, the disease-specific estimates represent the best available evidence to date, derived from among the most comprehensive systematic reviews available in the neurosurgical literature. Additionally, surgical provider data arose from the most far-reaching effort yet to quantify the worldwide population of neurosurgeons, and involved a partnering effort with many active international neurosurgery initiatives such as the WFNS.

Collaborators

We would like to thank the following individuals for their dedication and contribution to identifying the global neurosurgical deficit. Collaborators are listed in alphabetical order:

Amos O. Adeleye, MBBS, Amit Agrawal, MCh, Blake C. Alkire, MD, MPH, Julia R. Amundson, BS, Hildo Azevedo-Filho, MD, PhD, Ronnie E. Baticulon, MD, Joseph S. Bell, PhD, Erica Bisson, MD, MPH, Kamila M. Bond, BA, Vivek P. Buch, MD, Matthew C. Davis, MD, Robert Dempsey, MD, Serena Faruque, MS, PhD, Laurence Glancz, MBBS, BSc, William B. Gormley, MD, MPH, MBA, Saksham Gupta, BA, Michael M. Haglund, MD, PhD, MACM, Roger Härtl, MD, Joshua D. Hughes, MD, Ya Ching Hung, MD, Yoko Kato, MD, PhD, Robert M. Koffie, MD, PhD, Ramesh Kumar, MD, Jacob R. Lepard, MD, Jaims Lim, BS, Muhammad Raji Mahmud, MD, John G. Meara, MD, DMD, MBA, Rania A. Mekary, MSc, PhD, Basant K Misra, MD, Jacques J. Morcos, MD, Swagoto Mukhopadhyay, MD, Brian V. Nahed, MD, MSc, Enrique Osorio-Fonseca, MD, Sophie Peeters, BS, Maria Punchak, MSc, Mahmood M. Qureshi, MD, Christian Lopez Ramos, BS, Vijay Ravindra, MD, MSPH, Faith C. Robertson, BS, Jeffrey V. Rosenfeld, MD, Gail Rosseau, MD, Andrés M. Rubiano, MD, Sonal Sachdev, MD, Steven S. Senglaub, MS, Salman Y. Sharif, MD, Mark G. Shrime, MD, MPH, PhD, Kerry A. Vaughan, MD, Eka J. Wahjoepramono, MD, PhD, John C. Wellons III, MD, MSPH, and Ismaeel Yunusa, PharmD.

Online-Only Content

Supplemental material is available with the online version of the article.

• Appendix. https://thejns.org/doi/suppl/10.3171/2017.11.JNS171500.