An estimated 5 billion people worldwide are without access to surgical care.1 Barriers to surgical care can be framed chronologically as three delays—in seeking, reaching, and receiving care.2,3 The first two delays are challenges for patients in rural communities in low- and middle-income countries (LMICs), where the lack of access to health care services and limited transportation options routinely lead to late presentation of severe disease. Even if a patient reaches a health care facility, there is no guarantee that they will receive surgical treatment. Infrastructure and supply chain problems with essential medicines and basic supplies plague many hospitals in LMICs.1,4,5 World Health Organization surveys of hospitals in LMICs found that many of these facilities do not have access to reliable electricity or running water.1 Furthermore, many could not provide the “bellwether procedures” of a functioning surgical system, such as cesarean section (C-section), laparotomy, or open reduction of fractures.1
To address these barriers, Partners in Health and Zanmi Lasante (PIH/ZL) started to provide surgical care in central Haiti in 1996. These efforts evolved slowly but successfully, resulting in the construction of operating rooms (ORs) at three hospitals and a network of clinics for a catchment area of 3.3 million people.6 In 2008, an effort was made to increase surgical capacity with visiting surgeons and skilled support staff within the existing infrastructure of these three hospitals. In 2014, PIH/ZL sought to further scale-up surgical care and partnered with the Haitian Ministry of Public Health and Population to build the University Hospital of Mirebalais (UHM). Investments were made not just in training skilled surgical specialists, but also in infrastructure (ORs, surgical ward, and sterile processing) and in supply procurement. Surgical care for the catchment area also became centralized at UHM with a network of ambulances to transfer patients from the rural clinics and hospitals.
The authors conducted a study to evaluate PIH/ZL’s efforts to scale-up surgical services in central Haiti and overcome barriers to surgical care. The study used interrupted time-series analysis to assess changes in surgical volume. This article reports on the study design and findings and offers insights into a path forward.
The study’s primary outcome measure was the monthly total number of surgical operations, collected from OR logbooks at each PIH/ZL hospital in central Haiti. Surgical operations were defined as any procedure occurring in an OR involving the incision, excision, manipulation, or suturing of tissue, and requiring local, regional, or general anesthesia or sedation. Subgroup analyses for total numbers of C-section, laparotomy, and open reduction and fixation of fractures also were collected.
The interrupted time-series analysis was conducted using segmented linear regression and by dividing the time series into pre- and postintervention segments.7,8 The time periods each lasted six consecutive months (October–March) and were defined as follows:
- Time period one, 2007–2008: A preintervention or baseline time period for surgical care at the three PIH/ZL surgical facilities in central rural Haiti—Cange, Boucan-Carré, and Belladère
- Time period two, 2008–2009: A postintervention time period after initiating an effort to scale-up surgical care by increasing surgeons at the same three surgical facilities
- Time period three, 2014–2015: A postintervention time period after centralizing and scaling-up surgical services at UHM by increasing surgical staff, infrastructure, supply chain/procurement, and transportation
The segmented regression analysis assessed for changes in the level (step-change) and trend (slope-change) in the total number of operations between time periods.7 A change in the level constitutes a relatively abrupt effect of the intervention compared with the previous time segment, whereas a change in the slope represents a more gradual change in the total number of operations during a time segment. Observations in time-series data are often correlated, such that the total numbers of surgical operations in adjacent weeks are related.7,8 The study adjusted for first-order autocorrelation within the time series using the Prais–Winsten and Cochrane–Orcutt generalized least-squares estimators. Patient characteristics were summarized and compared between time periods. The null hypothesis was that the total number of operations per month was the same and constant across all time periods. The authors rejected the null hypothesis if statistical evidence indicated that the model parameters (level or slope) of the count of operations were statistically different at the p <0.05 level. All analyses were done using Stata/SE 12.1 statistical software. The study design was inspected by the human subjects committee at Harvard T.H. Chan School of Public Health, Boston, MA, and by PIH/ZL and was determined to be exempt from review as no identifying information about patients was recorded in this study.
In time period one (2007−2008), 33 surgeons (18 Haitian, 15 visiting surgeons) provided surgical care at facilities in Cange, Boucan-Carré, and Belladère, comprising a total of four ORs and 15 surgical ward beds. In time period two (2008−2009), 62 surgeons with increases in both the number of Haitian and visiting surgeons (31 Haitian, 31 visiting surgeons) operated at the same three sites with the same surgical ward capacity. In time period three (2014−2015), there were fewer total surgeons (53), although the ratio of local-to-visiting surgeons shifted (38 Haitian, 15 visiting surgeons). The surgeons in this group provided surgical care at one facility, UHM, with four active ORs and a 30-bed surgical ward.
In the preintervention time period one, a total of 1,019 operations were performed across the three PIH/ZL surgical facilities in central Haiti. During the postintervention time period two, after scaling-up surgical care with the addition of visiting surgeons within the existing infrastructure, 1,188 operations were performed. During postintervention time period three, after centralizing surgical services at UHM and scaling-up surgical care with additional Haitian surgical staff and fewer visiting surgeons as well as infrastructure, supply chain/procurement, and transportation, a total of 2,134 operations were performed (see Table 1 and Figure 1).
Table 1. Demographic data and operations by time period
The results of the segmented regression analyses are summarized in Table 2. In time period one, the baseline number of operations performed each month followed a significant and positive slope change with an average of 19.3 additional operations performed each month (p = 0.009). With attempts to scale-up surgical care through visiting surgeons in time period two, a significant downward step-change of 59.9 operations per month was detected when compared with time period one (p = 0.05), with no significant change in the slope (p = 0.48).
Table 2. Interrupted time-series regression analyses of total major operations
When surgical services were centralized to the new UHM, along with an increase in local rather than visiting surgical staff and increases in infrastructure, supply chain, and transportation, the total number of operations each month nearly doubled, with a significant upward step-change of 121.8 additional operations each month above the baseline trend of time period two (p = 0.001). No change occurred in the slope of the number of operations performed during time period three (p = 0.08).
The total number of operations is summarized for each time period (see Table 1). Among these operations, we performed a subgroup analysis of C-section, laparotomy, and open reduction and fixation of fracture. Figure 2 depicts trends in these procedures, with the most notable change between time periods two and three, resulting in a modest increase in the number of exploratory laparotomies, a threefold increase in the number of C-sections, and a disproportionately larger (tenfold) increase in the number of open reduction and fixation of fractures.
Figure 2. Comparison of counts of bellwether procedures by time period
A path forward
Scaling-up surgical care requires more than just the presence of additional surgeons. Investments in infrastructure, skilled surgical staff, supply chain management, and transportation are all critical to increasing surgical capacity. This study showed that when the number of surgeons in the Central Plateau was increased without addressing barriers responsible for delays in reaching patients or in the provision of surgical care, no corresponding increase in surgical volume occurred. When increases in surgeons shifted from higher numbers of visiting surgeons to higher numbers of local surgeons, and this change was paired with improved hospital infrastructure (ORs, surgical ward, sterilization equipment), supply chain/procurement, and transportation to the facility, surgical volume increased substantially. Nevertheless, such investments are sometimes considered to be expensive and often are rejected in favor of perceived lower-cost alternatives to providing care.
Across all time periods, the study data showed variation in surgical volume that is most likely attributable to seasonal changes in staffing. For example, a decrease occurs in November and December in all three time periods, as this time frame is when most local staff take their vacation. Changes also occurred in the trend, with decreases in surgical volume in February 2008 of time period one and March 2009 of time period two. Interestingly, these declines both corresponded with collaborative cleft lip/palate surgical missions the following month. Although the goal of these missions was to provide surgical care to children with cleft lip/palate, our data also suggests that these trips came at the expense of overall surgical volume. This situation may have arisen because surgeons and staff shifted their focus from operating to administrative preparation for the mission or may have been because the team wanted to ensure that the surgical ward would be able to accommodate patients during the mission.
Inadequate hospital infrastructure and supply chain problems may have limited the impact of additional visiting surgeons in time period two. As in the case of the mission trips, the number of beds in the surgery ward and the number of ORs proved particularly important. In the first two study periods, if the surgery wards were full, elective operations were canceled despite the presence of additional surgical staff or visiting surgeons. Increasing surgery ward capacity likely improved patient flow postoperatively and increased surgical volume despite no change in the number of ORs and a decrease in the number of visiting surgeons. Although the total number of ORs remained the same among the three study periods, centralizing them at one hospital also may have increased their use, as the more remote surgical sites were staffed less frequently.
It was initially important to provide care at all three remote sites in the Central Plateau because the roads were unpaved and routinely flooded during the rainy season, creating impassable barriers to patients from one region seeking care in another. Nonetheless, providing surgical care at multiple sites presented serious logistical challenges, leading to interruptions and delays in care as a result of power outages, equipment breakdown, supply stock-outs, and personnel reaching remote sites. When Route Nationale #3 was completed in 2011, connecting the provinces within central Haiti to one another with better roads and connecting the rural Central Plateau to Port-au-Prince, PIH/ZL was able to centralize surgical services in Mirebalais without substantially reducing access for patients in Cange, Boucan-Carré, and Belladère.
The study findings should not serve as a rejection of the principle of decentralization. Decentralizing health care delivery from urban centers to rural areas in LMICs is a well-supported strategy to increase access to care and guided PIH/ZL’s decision to build UHM in the rural Central Plateau instead of in Port-au-Prince. Rather, this study highlights the importance of actively remodeling health care networks to optimize efficiency along with parallel infrastructure development projects.
The increase in the total number of operations is a useful measure of surgical capacity, but looking at the trends in the bellwether procedures provides more information about the quality of the scale-up and the burden of disease in the catchment area. In this study, the most significant increases in surgical volume were related to women’s health and trauma. The substantial increase in the volume of C-sections and urgent and lifesaving procedures represents the greatest success of scaling-up surgical care in central Haiti and underscores the demand for these services in LMICs. New roads facilitated the centralization of surgical services, which contributed to a more efficient system and also reduced the delay in reaching care.
New roads likely contributed to reducing barriers in reaching care, but also brought new complications in the form of orthopaedic trauma. Assuming that the threefold increase in emergency obstetric care describes the effect of increased access to care, the tenfold increase in the number of fracture reductions is more than can be explained by increased access alone, and is likely the result of increased motor vehicle accidents on newly constructed roads. A well-supported surgical system has a critical role in addressing new road traffic-related trauma. Of course, the enormous burden of disease also highlights the importance of road safety management and other proactive interventions to reduce the traffic accidents that predictably accompany major road construction in LMICs.9
Lastly, the disproportionately smaller increase in exploratory laparotomies as compared with C-sections and open fracture reductions was surprising. Investments in infrastructure, such as reliable electricity and equipment maintenance, enabled use of new technology, such as computed tomography, and implementation of new procedures, such as endoscopy, which likely decreased the need for exploratory laparotomies. Other studies have demonstrated the role of ultrasound in reducing the incidence of exploratory laparotomies in Rwanda.10,11 Further studies can examine the potential value of less invasive endoscopy and imaging when compared with exploratory surgery in LMICs.
In conclusion, Haiti and other LMICs have a substantial unmet burden of surgical disease.1,12 This study should encourage continued investment in surgical infrastructure, procurement and supply chain management, and local skilled surgical specialists in resource-poor settings as the global community works to increase access to safe, timely, and affordable surgical care.
- Meara JG, Leather AJM, Hagander L, et al. Global Surgery 2030: Evidence and solutions for achieving health, welfare, and economic development. Lancet. 2015;386:569-624.
- Thaddeus S, Maine D. Too far to walk: Maternal mortality in context. Soc Sci Med. 1994;38(8):1091-1110.
- Pacagnella RC, Cecatti JG, Osis MJ, Souza JP. The role of delays in severe maternal morbidity and mortality: Expanding the conceptual framework. Reprod Health Matters. 2012;20(39):155-163.
- Kushner AL, Cherian MN, Noel L, Spiegel DA, Groth S, Etienne C. Addressing the Millennium Development Goals from a surgical perspective: Essential surgery and anesthesia in 8 low- and middle-income countries. Arch Surg. 2010;145(2):154-159.
- Wagenaar BH, Gimbel S, Hoek R, et al. Stock-outs of essential health products in Mozambique—longitudinal analyses from 2011 to 2013. Trop Med Int Health. 2014;19(7):791-801.
- Sullivan SR, Taylor HOB, Pauyo T, Steer ML. Surgeons’ dispatch from Cange, Haiti. N Engl J Med. 2010;362(7):e19.
- Wagner AK, Soumerai SB, Zhang F, Ross-Degnan D. Segmented regression analysis of interrupted time series studies in medication use research. J Clin Pharm Ther. 2002;27(4):299-309.
- Morgan OW, Griffiths C, Majeed A. Interrupted time-series analysis of regulations to reduce paracetamol (acetaminophen) poisoning. PLoS Med. 2007;4(4):e105.
- Moroz PJ, Spiegel DA. The World Health Organization’s action plan on the road traffic injury pandemic: Is there any action for orthopaedic trauma surgeons? J Orthop Trauma. 2014;28 Suppl 1:S11–14.
- Shah SP, Epino H, Bukhman G, et al. Impact of the introduction of ultrasound services in a limited resource setting: Rural Rwanda 2008. BMC Int Health Hum Rights. 2009;27(9):4.
- Groen RS, Leow JJ, Sadasivam V, Kushner AL. Review: Indications for ultrasound use in low- and middle-income countries. Trop Med Int Health. 2011;16(12):1525-1535.
- Weiser TG, Haynes AB, Molina G, et al. Estimate of the global volume of surgery in 2012: An assessment supporting improved health outcomes. Lancet. 2015;385 Suppl 2:S11.