General Surgical Services at an Urban Teaching Hospital in Mozambique

Abstract

Background

As surgery becomes incorporated into global health programs, it will be critical for clinicians to take into account already existing surgical care systems within low-income countries. In order to inform future efforts to expand the local system and systems in comparable regions of the developing world, we aimed to describe current patterns of surgical care at a major urban teaching hospital in Mozambique.

Methods

We performed a retrospective review of all general surgery patients treated between August 2012 and August 2013 at Hospital Central Maputo in Maputo, Mozambique. We reviewed emergency and elective surgical logbooks, inpatient discharge records, and death records to report case volume, disease etiology, and mortality.

Results

There were 1,598 operations (910 emergency, 688 elective) and 2,606 patient discharges during our study period. The most common emergent surgeries were for non-trauma laparotomy (22%) followed by all trauma procedures (18%), while the most common elective surgery was hernia repair (31%). The majority of lower extremity amputations were above knee (69%). The most common diagnostic categories for inpatients were infectious (31%), trauma (18%), hernia (12%), neoplasm (10%), and appendicitis (5%). The mortality rate was 5.6% (146 deaths), approximately half of which were related to sepsis.

Conclusions

Our data demonstrate the general surgery caseload of a large, academic, urban training and referral center in Mozambique. We describe resource limitations that impact operative capacity, trauma care, and management of amputations and cancer. These findings highlight challenges that are applicable to a broad range of global surgery efforts.

1 Introduction

Surgery has long been considered the “neglected stepchild” of global health. However, over the past decade, there has been an accumulation of evidence to suggest that surgical disease accounts for a substantial proportion of the global burden of disease and that the provision of surgical services in the developing world can be done both safely and cost-effectively.¹ As surgery is incorporated into global health programs, it will be critical to ensure that newly developed global health programs take into account the already existing systems and services in each country.² While surgical care in low- and middle-income countries is limited, most countries have an already established system that has evolved within the context of the local health care system. Developing a thorough understanding of each country's system of surgical care delivery is therefore a prerequisite to affecting meaningful and sustainable improvements.

Mozambique, a low-income country in Southeast Africa, offers a compelling case study. The country has long suffered from a severe shortage of physicians—particularly surgeons. In a country with a population of over 25 million, there are an estimated 116 surgical specialists, 38 of whom are general surgeons. Over half of the surgeons providing care in Mozambique are foreign doctors. Mozambique also relies heavily on mid-level surgical providers, ‘tecnicos de cirurgia,’ to deliver general surgical care. Throughout the country, there are an estimated 63 ‘tecnicos,’ 12 of who work in the outskirts of Maputo and 51 of who work in more remote regions of the country.

Hospital Central de Maputo (HCM) is a 1500-bed tertiary care center, the largest of three public hospitals in Maputo, and serves an estimated population of 3 million, including residents of Maputo as well as those of the surrounding region in southern Mozambique. HCM is the only hospital in the country equipped to handle advanced operations, thereby serving as the quarternary referral center for the entire country. However, it is equipped with only eleven operating rooms (and fourteen operating tables) to service all surgical specialties. Meanwhile, HCM is also the training hospital for Eduardo Mondlane University, the country's primary medical school, offering several surgical residency programs, including subspecialties such as plastic surgery, orthopedic surgery, and pediatric surgery. Surgery education has expanded rapidly in the past few years. Currently, up to 8 candidates are accepted annually into the general surgery residency program at HCM/Eduardo Mondlane University. The minimum number of required cases to graduate is 1000 (400 as primary surgeon). Opportunities exist for short-term training abroad, but all Mozambicans reportedly return to practice in their country. In addition, surgeons trained at HCM serve as educators for the ‘tecnicos,’ who end up working primarily in hospitals outside of Maputo, where human and material resources are even more limited.³

Local surgeons and surgical trainees have expressed increased interest in participating in research projects in order to better understand current patterns of care and ultimately to improve the quality of care delivered. However, to date, these efforts have been limited by the lack of reliable data on the current case mix and volume of the hospital. In order for HCM—and, by extension, all of Mozambique—to continue its trajectory toward improved surgical care, a better understanding of current patterns of care delivery at HCM is needed. In partnership with local surgeons, we retrospectively evaluated the general surgery caseload over a one-year period. By systematically describing the existing patterns of general surgical care, we aimed to inform efforts to expand and improve surgical training locally as well as in comparable regions in the developing world.

2 Methods

We performed a retrospective review of all general surgery patients treated from August 2012 to August 2013 at Hospital Central Maputo. In order to describe patients that were managed both operatively and non-operatively, we obtained data from two sources. First, we reviewed the general surgery logbooks, which were maintained for both emergent and elective cases performed during our study period. Second, in order to include patients that did not undergo an operation, we reviewed ward discharge records and hospital death records. Pediatric patients were excluded unless their operation was performed as an emergency procedure by general surgery staff and was recorded in the general surgery emergency operating room logbook.

To describe the case volume, we divided operations into emergent and elective categories based on the location where the procedure was performed. Data for emergent cases was obtained from logbooks located in the operating rooms near the emergency room. Data for elective cases was obtained from logbooks located in the operating rooms in the main hospital. Within each category, we further subdivided the cases into categories relating to operation type and organ system. Initial categories were developed a priori based on our expectations and anecdotal experience in the hospital, but were then refined based on empirical evidence. Those that did not fit into one of our categories were labeled as ‘other’. Each step of the data review was performed in partnership with local surgeons and academic partners.

We performed a similar analysis for the diagnoses obtained from the inpatient discharge records and death records within the general surgery department. For all analyses, we calculated the mean age and the gender distribution for each category. Unpaired t-test and Fisher's exact test were used to compare continuous variables and proportions, respectively.

3 Results

During our study period, there were a total of 1,598 operations, 2,606 patient discharges, and 146 deaths (144 post-operative, 2 intraoperative) within the general surgery department.

3.1 Major Operative Cases

Of the 1,598 major surgical procedures, 910 were emergent and 688 were elective (Table 1). No operations were performed laparoscopically. The overall mean age for patients undergoing an operation was 40 years, and 58% of the patients were male.

Table 1.

General Surgery Operative Cases

	Total n(%)	Mean Age ± SD	Male n(%)
Emergency Operations	910(100)	37.5 ± 17.0	586(64)
Laparotomy (Non-Trauma)	203(22)	36.2 ± 15.0	99 (49)
Gastrointestinal	157
Gynecologic	46
Trauma	163(18)	30.4 ± 14.0	113(69)
Wound Closure/Debridement	39
Splenectomy/Splenorraphy	31
Bowel Trauma	31
Other	27
Negative Laparotomy	20
Liver Trauma	15
Appendectomy	138(15)	28.6 ± 11.6	87(63)
Hernia Repair	114(13)	35.8 ± 13.4	103(90)
Inguinal	103
Ventral	10
Richter	1
Amputation	109(12)	58.0 ± 15.0	68(62)
Above Knee	87
Below Ankle	13
Upper Extremity	3
Revision	4
Below Knee (Above Ankle)	2
Incision and Drainage	101(11)	36.0 ± 15.2	62(61)
Debridement (Non-Trauma)	52(6)	50.5 ± 18.4	36(69)
Other	12(1)	43.5 ± 19.0	7(58)
Elective Operations	688(100)	42.4 ± 16.3	347(50)
Hernia Repair	211(31)	44.9 ± 16.8	137(65)
Inguinal	142
Ventral	58
Hydrocelectomy	9
Femoral	2
Laparotomy	125(18)	40.5 ± 15.3	66(53)
Stoma Formation	26
Stoma Closure	21
Exploratory Laparotomy	21
Cholecystectomy	17
Splenectomy	12
Bowel Resection	18
Other	10
Anorectal	111(16)	38.9 ± 13.4	63(57)
Hemorrhoidectomy	64
Fistulotomy	25
Sphincterotomy	15
Other	7
Breast Surgery	84(12)	37.2 ± 17.7	9(11)
Excisional Biopsy	41
Mastectomy	35
Other	8
Amputation	63(9)	52.3 ± 16.7	39(62)
Above Knee	33
Below Ankle	13
Revision	10
Upper Extremity	5
Below Knee (Above Ankle)	2
Thyroid Procedures	37(5)	42.1 ± 12.6	3(7.9)
Excisional Biopsy	31(5)	38.5 ± 15.3	16(52)
Wound Debridement	22(3)	46.1 ±15.8	13(59)
Other	4(1)	40.8 ±16.8	3(75)

Emergency operations

Emergency operations comprised 57% of general surgery operations. Common reasons for emergent surgery were non-trauma laparotomy (22%), trauma (18%), appendectomy (15%), hernia repair (13%), amputation (12%), and incision and drainage (11%). Non-trauma laparotomies included bowel resections and closure of gastrointestinal perforations (n=55), salpingectomies for ruptured ectopic pregnancy (n=37), lysis of adhesions (n=36), negative laparotomies (n=20), volvulus reduction and/or resection (n=20), intra-abdominal tumor excisions and/or biopsies (n=9), and ovarian or uterine mass excisions (n=9).

Of the 163 trauma cases, the most common indication was blunt abdominal trauma (33%), followed by stab wound (17%) and gunshot wound (9%). Of the 31 operative cases involving splenic trauma, 87% resulted in splenectomy and 13% underwent splenorraphy. A total of 121 laparotomies were performed for trauma, 20 of which (17%) were negative laparotomies. The vast majority (80%) of the negative laparotomies were performed for the indication of blunt abdominal trauma. Mean age for patients undergoing operations for trauma (30 years) was significantly lower than the mean age for all patients undergoing emergent procedures (38 years; p<0.001).

During the study period, 109 emergent limb amputations were performed for reasons other than acute traumatic injury including wet gangrene and diabetic disease. Mean age for these patients undergoing amputations (58 years) was significantly higher than the mean age for all patient undergoing emergent procedures (38 years; p<0.001). Lower extremity amputations (n=106) were usually done above the knee (n=87). Only 2 emergent below knee (above ankle) amputations were recorded in the year of study.

Elective operations

Of 688 elective operations, the most commonly performed were hernia repair (29%), laparotomy (18%), anorectal surgery (16%), breast surgery (12%), and limb amputations (9%). Most of the hernia repairs were for inguinal hernias (n=142, 70%), usually on males (63% of all elective hernias). Laparotomy procedures were most commonly performed for the creation or closure of an ostomy (n=47, 38%), for abdominal exploration, typically to obtain tissue samples for biopsy (n=21, 17%), or for an intestinal bypass (n=18, 14%). The most common indication for elective laparotomy was colorectal cancer.

Hemorrhoids were the most common indication for anorectal surgery. Breast surgery was most commonly performed for excisional biopsies (n=41, 49%). The next most common breast operation was mastectomy (n=35, 42%), all of which were done for cancer. Of note, there were no sentinel lymph node biopsies performed.

Elective amputations were most commonly associated with dry gangrene and diabetic neuropathy. Once again, the majority were lower extremity amputations (n=58, 92%) and only 2 of those were below knee (above ankle). Revisions of prior lower extremity amputations accounted for 16% of elective amputations.

Thyroid procedures accounted for 5% (n=37) of the elective operations. There were 30 partial thyroidectomies, 6 total thyroidectomies, and 1 enucleation of a thyroid nodule. Indications for thyroidectomy included goiter (19 multinodular goiters, 5 unspecified goiters, and 4 nontoxic diffuse goiter), and cancer (6 follicular carcinomas, 2 papillary carcinomas). Excisional biopsies of masses in the upper and lower extremities and abdomen accounted for 5% (n=31) of elective operations. There were 22 non-emergent wound debridement procedures. Finally, there was one case of a femoral artery repair and three cases in which the procedure could not be determined due to illegible records.

3.2 Ward Discharge Data

The most common diagnostic categories for inpatient discharge were infectious (31%), trauma (18%), hernia (12%), neoplastic (10%), and appendicitis (5%) (Table 2). Common infectious diagnoses included abscesses, erysipelas, pressure ulcers, and necrotizing fasciitis. Common traumatic diagnoses included polytrauma, blunt abdominal trauma, stab wounds, and gunshot wounds. A disproportionately high percentage of males was noted in the trauma (74%; p<0.001) and hernia (74%; p<0.001) categories. Mean age of patients admitted with appendicitis (30 years) and trauma (33 years) was significantly lower than the population as a whole (40 years) (p<0.001). The other diagnostic category included the following diagnoses: 25 illegible or unknown, 10 unspecified hemoperitoneum, 4 mastitis, 4 unspecified hematomas, 3 varicose veins, 2 anemia, 2 diabetes, 2 foreign body, 2 adult imperforate anus, 2 pulmonary embolism, 2 thrombophlebitis, and 1 each of ascites, costochondritis, diverticulitis, diverticulosis, dysphagia, elephantiasis, esophageal stenosis, esophagitis, fecal impaction, unspecified foot lesion, gastristis, gynecomastia, unspecified hemorrhage, hidradenitis, lymphangitis, paraplegia, pneumothorax, post-hysterectomy, post-laparotomy, post-op fistula unspecified, and tenosynovitis.

Table 2.

General Surgery Ward Discharges Based on Diagnostic Category

Diagnostic Category	Number	Percent	Average Age	(Standard Deviation)	Male	(%)
Total	2606	100%	40.4	16.9	1482	58
Infectious	805	31%	44.3	18.1	430	53
Trauma	458	18%	33.0	13.4	339	74
Hernia	310	12%	42.0	16.1	237	76
Neoplastic	256	10%	41.6	16.1	74	29
Appendicitis	128	5%	29.5	11.0	79	62
Anorectal	108	4%	39.9	12.4	74	69
Intestinal Obstruction	85	3%	39.9	17.5	46	54
Extremity Gangrene	64	2%	57.5	17.2	41	64
Biliary	38	1%	45.2	13.9	8	21
Goiter	30	1%	41.9	9.7	1	3
Ostomy-Related	30	1%	37.0	15.4	21	70
Peritonitis	29	1%	28.3	9.9	17	59
Ruptured Ectopic	28	1%	30.0	5.8	0	0
Deep Vein Thrombosis	26	1%	41.7	12.8	5	19
Gastrointestinal Bleeding	21	1%	47.0	21.7	10	48
Diabetic Foot	20	1%	55.5	15.1	12	60
Pancreatitis	20	1%	40.0	9.8	16	80
Hypersplenism	17	1%	37.2	11.1	6	35
Lymphedema	16	1%	50.6	17.1	8	50
Abdominal Pain	15	1%	39.7	18.0	9	56
Wound Dehiscence	14	1%	45.0	18.2	8	57
Enterocutaneous Fistula	10	0%	31.2	11.8	2	20
Other	78	3%	40.6	19.2	39	50

3.3 Mortality Data

There were 146 deaths recorded (144 post-operative, 2 intra-operative), yielding a mortality rate for the general surgery ward of 5.6%. Both intra-operative deaths occurred as a result of uncontrolled hemorrhage. Of the patients that died during the hospital stay, the mean age was 42 years, and 49% were male. The majority of deaths were attributed to an infectious diagnosis and 53% had sepsis listed as a cause of death. Neoplasm was recorded as a comorbidity in 14% of deaths, and HIV was recorded as a comorbidity in 14% of deaths.

4 Discussion

In partnership with local surgeons, we describe the current patterns of general surgical care delivery in a major urban teaching hospital in Maputo, Mozambique. The number of general surgery cases performed per annum in this 1500-bed hospital is less than that performed at public urban hospitals in the United States with much smaller bed capacity, and it reflects limitations in operative capacity including number of operating rooms and availability of staffing and equipment. Efforts to increase the surgical workforce with expansion of the surgical residency programs should be accompanied by an expansion in infrastructure and operative capacity to provide trainees with adequate experience.

It is also important that the training and experience be contextually relevant. There are limited data available on the case mix and volume for general surgery operations in sub-Saharan Africa. In this study, we found a notably high proportion of operations performed emergently (57%), frequently for trauma and incarcerated hernias. In addition, we noted that the general surgery service managed some emergency gynecologic pathologies including ruptured ectopic pregnancies. We also noted a high prevalence of elective operations ranging from hernia repairs to laparotomies, as well as anorectal and breast operations. These findings provide a useful foundation for understanding existing surgical care patterns in low-income countries and can be used to highlight existing strengths as well as important shortcomings that need to be addressed.

Emergency and trauma care are a major priority for Mozambique. We found that the majority of operative cases were performed emergently, and a large percentage was for trauma. In particular, a large number of traumatic splenectomies and negative laparotomies were noted. Improvements in critical care and imaging, including increased availability and quality of ultrasound and computed tomography, could facilitate non-operative management of trauma and help address limited operative capacity. For example, in response to an increased awareness about trauma and road accidents, ^4,5,6,7,8 the Ministry of Health (MISAU) has established a call center and funded the first ambulances for a pre-hospital trauma system. In addition, MISAU has partnered with the Primary Trauma Care Foundation (PTCF) to offer contextually-appropriate trauma courses. Furthermore, a trauma registry has been establish at HCM in partnership with McGill University.⁹ In the general surgery department, there is interest in training trauma and critical care surgeons in order to support the emerging trauma care system in Mozambique.

We evaluated records over an entire one-year period, but found essentially no advanced surgeries, such as laparoscopic operations, pancreatectomies, or liver resections. Low- and middle-income countries (LMICs) have been slow to adopt these advanced operations, likely due to the significant resources required and the fact that these operations have a smaller public health impact and are less cost-effective than more routine operations.^10,11 Nevertheless, successful laparoscopic programs have been implemented in other developing countries, such as Tanzania, Nigeria, and Mongolia. These programs have relied on skills training from visiting specialists and cost-reduction strategies made by local surgeons.^12,13,14,15 While some equipment was donated in the past to both the general surgery and the obstetrics departments at HCM, the donations were not accompanied by a formalized training program, which greatly limited adoption of the technology.

Our data reflect many examples of how resource limitations affect surgical care, most notably in relation to amputations and cancer care. The vast majority of lower extremity amputations are performed to treat diabetic complications and since this disease is frequently progressive, the operation of choice for these patients is almost always an above-knee amputation. Whereas a below-knee amputation results in better mobility, improved quality of life, and even longer life expectancy,¹⁶ the above-knee operation is frequently chosen simply to avoid the likely need for a subsequent operation. Improvements in the availability of diabetic care, vascular surgery, and rehabilitation services could help to prevent and limit the morbidity of amputations. Similarly, cancer care at HCM is limited by resource availability. While neoplasms accounted for 9% of all surgical discharge diagnoses, we found few examples of oncologic resections in our data. Instead, patients with cancer often underwent either biopsy or diversion, possibly due to the delayed presentation and already advanced disease. Also, breast cancer patients were typically treated with mastectomy alone without a sentinel lymph node biopsy, possibly due to the fact that radiation therapy is not available in the public healthcare system in Mozambique.

There are several limitations to this retrospective study. First, our data were obtained from a referral and teaching hospital, and it does not necessarily represent the spectrum of disease seen by general surgeons in district hospitals. The primary goal of this study, however, was to identify the current surgical capacity of a large teaching hospital in a low-income country, independent of the care delivered in the surrounding district hospitals. Second, we only evaluated data for one year; however, this period was representative of the average case volume. Third, we obtained data from surgery logbooks, which were handwritten, often illegibly and without a standardized template, which resulted in frequently missing data fields. However, with the help of local surgeons, we were able to accumulate a large number of data points, which we believe are reflective of the hospital's typical annual caseload. Finally, aside from describing the operations performed, we were unable to measure or record postoperative patient outcomes during the study period. While clinical registries are expensive and challenging to maintain, there are efforts on the national level in Mozambique to more reliably capture postoperative outcomes to encourage nationwide quality improvement efforts.

By carefully evaluating patient records over a one-year period, we have summarized the typical general surgery caseload for a large, academic, urban referral center in a low-income country. Our data demonstrate a well-established surgical program providing important care to a large patient population. However, we also found significant shortcomings that may be improved by carefully implementing well-designed, partnered interventions. We believe these findings highlight challenges that are applicable to a broad range of global health efforts.