Face-to-face interviews versus Internet surveys: Comparison of two data collection methods in the Rome foundation global epidemiology study: Implications for population-based research.

BACKGROUND AND AIMS: The Rome Foundation Global Epidemiology Study (RFGES) assessed the prevalence, burden, and associated factors of Disorders of Gut-Brain Interaction (DGBI) in 33 countries around the world. Achieving worldwide sampling necessitated use of two different surveying methods: In-person household interviews (9 countries) and Internet surveys (26 countries). Two countries, China and Turkey, were surveyed with both methods. This paper examines the differences in the survey results with the two methods, as well as likely reasons for those differences. METHODS: The two RFGES survey methods are described in detail, and differences in DGBI findings summarized for household versus Internet surveys globally, and in more detail for China and Turkey. Logistic regression analysis was used to elucidate factors contributing to these differences. RESULTS: Overall, DGBI were only half as prevalent when assessed with household vs Internet surveys. Similar patterns of methodology-related DGBI differences were seen within both China and Turkey, but prevalence differences between the survey methods were dramatically larger in Turkey. No clear reasons for outcome differences by survey method were identified, although greater relative reduction in bowel and anorectal versus upper gastrointestinal disorders when household versus Internet surveying was used suggests an inhibiting influence of social sensitivity. CONCLUSIONS: The findings strongly indicate that besides affecting data quality, manpower needs and data collection time and costs, the choice of survey method is a substantial determinant of symptom reporting and DGBI prevalence outcomes. This has important implications for future DGBI research and epidemiological research more broadly.

A comparative cost analysis of two screening strategies for colorectal cancer in Lynch Syndrome in a South African tertiary hospital.

AIM: Lynch Syndrome (LS) individuals have a 25-75% lifetime risk of developing colorectal cancer. Colonoscopy screening decreases this risk. This study compared the cost of Strategy 1: screening colonoscopy for 1st degree relatives of patients that met the Revised Bethesda Criteria (i.e., probands) to Strategy 2: screening colonoscopy for 1st degree relatives of probands with genetic mutations for Lynch Syndrome based in a resource-constrained health care system. METHOD: A comparative, health care provider perspective, cost analysis was conducted at a tertiary hospital, using a micro-costing, ingredient approach. Forty probands that underwent genetic testing between November 01, 2014 and October 30, 2015 and their first-degree relatives were costed according to Strategy 1 and Strategy 2. Unit costs of colonoscopy and genetic testing were estimated and used to calculate and compare the total costs per strategy in South African rand (R) converted to UK pounds (£). Sensitivity analyses were performed on colonoscopy adherence, relatives' positivity, and variable discount rates. RESULTS: The cost for Strategy 1 amounted to £653 344/R6 161 035 compared to £49 327/R 465 155 for Strategy 2 (Discount rate 3%; Adherence 75%; and Positivity rate of relatives 45%). Base case analysis indicated a difference of 92% less in the total cost for Strategy 2 compared to Strategy 1. Sensitivity analyses showed that the difference in cost between the two strategies was not sensitive to variations in adherence, positivity or discount rates. CONCLUSION: Colonoscopy screening for LS and at-risk family members was tenfold less costly when combined with genetic analysis. The logistics of rolling out this strategy nationally should be investigated.

Prognosis in acute pancreatitis associated with HIV infection.

BACKGROUND: This prospective study compared scoring systems in predicting adverse outcomes in HIV associated acute pancreatitis (HIV+ve AP) METHODS: Systemic inflammatory response syndrome (SIRS), Glasgow criteria, C-reactive protein (CRP), bedside index of severity in acute pancreatitis (BISAP) and APACHE II scores using standard cut-off values were used to predict the endpoint of moderate and severe disease in HIV-ve and HIV+ve patients and in CD4 counts above and below 200 cells/mm3. RESULTS: Ninety (38%) of 238 patients with AP were HIV+ve. Fifteen had organ failure, 33 local complications and 12 patients died. Advanced age was not associated with severe disease. The APACHE II was the best predictor of severe disease in HIV+ve (AUC 0.88) and HIV-ve patients (AUC 0.81) and CRP was the poorest predictor (AUC 0.59) in HIV+ve patients. In HIV+ve patients with CD4 counts greater and less than 200 cells/mm3 the Glasgow and APACHE II scores were the best prognosticators (AUC > 0.8) and BISAP in patients with CD4 > 200 cells/mm3 (AUC 0.90). CONCLUSION: The APACHE II score was most effective irrespective of HIV status whereas the BISAP scores was better in CD4 > 200 cells/mm3.

Worldwide Prevalence and Burden of Functional Gastrointestinal Disorders, Results of Rome Foundation Global Study.

BACKGROUND & AIMS: Although functional gastrointestinal disorders (FGIDs), now called disorders of gut-brain interaction, have major economic effects on health care systems and adversely affect quality of life, little is known about their global prevalence and distribution. We investigated the prevalence of and factors associated with 22 FGIDs, in 33 countries on 6 continents. METHODS: Data were collected via the Internet in 24 countries, personal interviews in 7 countries, and both in 2 countries, using the Rome IV diagnostic questionnaire, Rome III irritable bowel syndrome questions, and 80 items to identify variables associated with FGIDs. Data collection methods differed for Internet and household groups, so data analyses were conducted and reported separately. RESULTS: Among the 73,076 adult respondents (49.5% women), diagnostic criteria were met for at least 1 FGID by 40.3% persons who completed the Internet surveys (95% confidence interval [CI], 39.9-40.7) and 20.7% of persons who completed the household surveys (95% CI, 20.2-21.3). FGIDs were more prevalent among women than men, based on responses to the Internet survey (odds ratio, 1.7; 95% CI, 1.6-1.7) and household survey (odds ratio, 1.3; 95% CI, 1.3-1.4). FGIDs were associated with lower quality of life and more frequent doctor visits. Proportions of subjects with irritable bowel syndrome were lower when the Rome IV criteria were used, compared with the Rome III criteria, in the Internet survey (4.1% vs 10.1%) and household survey (1.5% vs 3.5%). CONCLUSIONS: In a large-scale multinational study, we found that more than 40% of persons worldwide have FGIDs, which affect quality of life and health care use. Although the absolute prevalence was higher among Internet respondents, similar trends and relative distributions were found in people who completed Internet vs personal interviews.

Gender Differences in Oesophageal Squamous Cell Carcinoma in a South African Tertiary Hospital.

(1) Oesophageal squamous cell carcinoma is common in Africa and has a male preponderance. The gender-based differences in clinical presentation and risk factor exposure are poorly studied in the African context. Our aim was to compare males and females with this disease. We analyzed the differences in clinical features and risk factor exposure between males and females with oesophageal cancer. (2) Data from patients presenting to a tertiary hospital in South Africa with oesophageal squamous cell carcinoma were analyzed. Data collected included patient demographics, clinical presentation, pathology and risk factor exposure. (3) Three hundred and sixty three patients were included in the study. The male to female ratio was 1.4:1. The mean age was 66 years for females and 61 years for males (p < 0.0001). A significantly larger percentage of males were underweight compared to females (60% vs. 32%, p < 0.001). There were no differences between the genders with regards to performance status, dysphagia grade and duration and tumor length, location and degree of differentiation. There were significant differences between risk factor exposure between the two genders. Smoking and alcohol consumption was an association in more than 70% of males but in less than 10% of females There was no difference survival. (4) Female patients with oesophageal squamous cell carcinoma (OSCC) are older and have a higher body mass index (BMI) than their male counterparts. Traditionally purported risk factors of smoking and alcohol consumption are infrequent associations with OSCC in female patients and other environmental risk factors may be more relevant in this gender.

Personal Protective Equipment for Endoscopy in Low-Resource Settings During the COVID-19 Pandemic: Guidance From the World Gastroenterology Organisation.

Performance of endoscopic procedures is associated with a risk of infection from COVID-19. This risk can be reduced by the use of personal protective equipment (PPE). However, shortage of PPE has emerged as an important issue in managing the pandemic in both traditionally high and low-resource areas. A group of clinicians and researchers from thirteen countries representing low, middle, and high-income areas has developed recommendations for optimal utilization of PPE before, during, and after gastrointestinal endoscopy with particular reference to low-resource situations. We determined that there is limited flexibility with regard to the utilization of PPE between ideal and low-resource settings. Some compromises are possible, especially with regard to PPE use, during endoscopic procedures. We have, therefore, also stressed the need to prevent transmission of COVID-19 by measures other than PPE and to conserve PPE by reduction of patient volume, limiting procedures to urgent or emergent, and reducing the number of staff and trainees involved in procedures. This guidance aims to optimize utilization of PPE and protection of health care providers.

Investigation on the hereditary basis of colorectal cancers in an African population with frequent early onset cases.

BACKGROUND: Approximately 25% of colorectal cancer patients in sub-Saharan Africa are younger than 40 years, and hereditary factors may contribute. We investigated the frequency and patterns of inherited colorectal cancer among black Zimbabweans. METHODS: A population-based cross-sectional study of ninety individuals with a new diagnosis of colorectal cancer was carried out in Harare, Zimbabwe between November 2012 and December 2015. Phenotypic data was obtained using interviewer administered questionnaires, and reviewing clinical and pathology data. Cases were screened for mismatch repair deficiency by immunohistochemistry and/or microsatellite instability testing, and for MLH1, MSH2 and EPCAM deletions using multiplex ligation-dependent probe amplification. Next generation sequencing using a 16-gene panel was performed for cases with phenotypic features consistent with familial colorectal cancer. Variants were assessed for pathogenicity using the mean allele frequency, phenotypic features and searching online databases. RESULTS: Three Lynch syndrome cases were identified: MSH2 c.2634G>A pathogenic mutation, c.(1896+1_1897-1)_(*193_?)del , and one fulfilling the Amsterdam criteria, with MLH1 and PMS2 deficiency, but no identifiable pathogenic mutation. Two other cases had a strong family history of cancers, but the exact syndrome was not identified. The prevalence of Lynch syndrome was 3·3% (95% CI 0·7-9·4), and that of familial colorectal cancer was 5·6% (95% CI, 1·8-12·5). CONCLUSIONS: Identifying cases of inherited colorectal cancer in sub-Saharan Africa is feasible, and our findings can inform screening guidelines appropriate to this setting.

A case-control study of risk factors for colorectal cancer in an African population.

The interplay between hereditary and environmental factors in the causation of colorectal cancer in sub-Saharan Africa is poorly understood. We carried out a community based case-control study to identify the risk factors associated with colorectal cancer in Zimbabwe. We recruited 101 cases of colorectal cancer and 202 controls, matched for age, sex and domicile. Potential risk factors including family history, socioeconomic status, urbanization, diabetes mellitus and previous schistosomiasis were evaluated. Conditional logistic regression was used to estimate the odds ratios associated with the different factors. Cases were more likely to have a tertiary education (32.7 vs. 13.4%, P<0.001) and a higher income (18.8 vs. 6.9%, P=0.002). After multivariate analysis, diabetes mellitus [odds ratio (OR): 5.3; 95% confidence interval (CI): 1.4-19.9; P=0.012], previous urban domicile (OR: 2.8; 95% CI: 1.0-7.8; P=0.042), previous schistosomiasis (OR: 2.4; 95% CI: 1.4-4.2; P=0.001) and cancer in a first-degree relative (OR: 2.4; 95% CI: 1.2-4.8; P=0.018) were associated independently with colorectal cancer. Our findings suggest that family history, diabetes mellitus, previous schistosomiasis and approximation to a western lifestyle are the predominant associations with colorectal cancer in Africans. This offers opportunities for targeted prevention and hypothesis-driven research into the aetiology of colorectal cancer in this population.

Dietary patterns and colorectal cancer risk in Zimbabwe: A population based case-control study.

BACKGROUND: The rising incidence of colorectal cancer in sub-Saharan Africa may be partly caused by changing dietary patterns. We sought to establish the association between dietary patterns and colorectal cancer in Zimbabwe. METHODS: One hundred colorectal cancer cases and 200 community-based controls were recruited. Data were collected using a food frequency questionnaire, and dietary patterns derived by principal component analysis. Generalised linear and logistic regression models were used to assess the associations between dietary patterns, participant characteristics and colorectal cancer. RESULTS: Three main dietary patterns were identified: traditional African, urbanised and processed food. The traditional African diet appeared protective against colorectal cancer (Odds Ratio (OR) 0.35; 95% Confidence Interval (CI), 0.21 - 0.58), which had no association with the urban (OR 0.68; 95% CI, 0.43-1.08), or processed food (OR 0.91; 0.58-1.41) patterns. The traditional African diet was associated with rural domicile, (OR 1.26; 95% CI, 1.00-1.59), and a low income (OR1.48; 95% CI, 1.06-2.08). The urbanised diet was associated with urban domicile (OR 1.70; 95% CI, 1.38-2.10), secondary (OR 1.30; 95% CI, 1.07-1.59) or tertiary education (OR 1.48; 95% CI, 1.11-1.97), and monthly incomes of $201-500 (OR 1.30; 95% CI, 1.05-1.62), and the processed food pattern with tertiary education (OR 1.42; 95% CI, 1.05-1.92), and income >$1000/month (OR 1.48; 95% CI, 1.02-2.15). CONCLUSION: A shift away from protective, traditional African dietary patterns may partly explain the rising incidence of colorectal cancer in sub-Saharan Africa.

Pancreatitis in a high HIV prevalence environment.

BACKGROUND: Acute pancreatitis is common in HIV-positive individuals in reports from regions with a low incidence of HIV infection. This association has not been reported in areas with a high incidence of HIV infection. OBJECTIVE: To examine the prevalence and outcomes of HIV-associated acute pancreatitis in a high HIV prevalence environment, and trends over the period May 2001 - November 2010. METHODS: The records of patients admitted with acute pancreatitis from 2001 to 2010 were reviewed, looking for HIV status, CD4 counts and medications at presentation. The Glasgow criteria, organ failure, local complications and mortality were assessed. RESULTS: One hundred and six (16.9%) of 627 patients admitted with acute pancreatitis during the study period were infected with HIV. Most were female (65.1%) and black African (91.5%). The serum amylase level was used to confirm acute pancreatitis in 50 patients, with a mean of 1 569 IU/L (range 375 - 5 769), and urinary amylase in 56 patients, with a mean of 4 083 IU/L (range 934 - 36 856). Alcohol was a less frequent cause of pancreatitis in the HIV-positive group than in patients who were HIV-negative (24.5% v. 68.3%), and the prevalence of gallstones as a cause was similar (23.6% v. 17.9%). Antiretroviral therapy was associated with pancreatitis in 35.8%, and 6 (5.7%) had abdominal malignancies. Sixteen (15.1%) had pancreatic necrosis, 20 (18.9%) had septic complications, and 6 (5.7%) died. CONCLUSIONS: HIV-associated acute pancreatitis was most frequent in females and black Africans and was associated with malignancy. Mortality was similar in HIV and non-HIV pancreatitis.

Resection of complex pancreatic injuries: Benchmarking postoperative complications using the Accordion classification.

AIM: To benchmark severity of complications using the Accordion Severity Grading System (ASGS) in patients undergoing operation for severe pancreatic injuries. METHODS: A prospective institutional database of 461 patients with pancreatic injuries treated from 1990 to 2015 was reviewed. One hundred and thirty patients with AAST grade 3, 4 or 5 pancreatic injuries underwent resection (pancreatoduodenectomy, n = 20, distal pancreatectomy, n = 110), including 30 who had an initial damage control laparotomy (DCL) and later definitive surgery. AAST injury grades, type of pancreatic resection, need for DCL and incidence and ASGS severity of complications were assessed. Uni- and multivariate logistic regression analysis was applied. RESULTS: Overall 238 complications occurred in 95 (73%) patients of which 73% were ASGS grades 3-6. Nineteen patients (14.6%) died. Patients more likely to have complications after pancreatic resection were older, had a revised trauma score (RTS) < 7.8, were shocked on admission, had grade 5 injuries of the head and neck of the pancreas with associated vascular and duodenal injuries, required a DCL, received a larger blood transfusion, had a pancreatoduodenectomy (PD) and repeat laparotomies. Applying univariate logistic regression analysis, mechanism of injury, RTS < 7.8, shock on admission, DCL, increasing AAST grade and type of pancreatic resection were significant variables for complications. Multivariate logistic regression analysis however showed that only age and type of pancreatic resection (PD) were significant. CONCLUSION: This ASGS-based study benchmarked postoperative morbidity after pancreatic resection for trauma. The detailed outcome analysis provided may serve as a reference for future institutional comparisons.

The shifting epidemiology of colorectal cancer in sub-Saharan Africa.

The perception that colorectal cancer is rare in sub-Saharan Africa is widely held; however, it is unclear whether this is due to poor epidemiological data or to lower disease rates. The quality of epidemiological data has somewhat improved, and there is an ongoing transition to western dietary and lifestyle practices associated with colorectal cancer. The impact of these changes on the incidence of colorectal cancer is not as evident as it is with other non-communicable diseases such as diabetes. In this Viewpoint, we discuss the epidemiology of colorectal cancer in sub-Saharan Africa. Colorectal cancer in this region frequently occurs at an early age, often with distinctive histological characteristics. We detail the crucial need for hypothesis-driven research on the risk factors for colorectal cancer in this population and identify key research gaps. Should colorectal cancer occur more frequently than assumed, then commensurate allocation of resources will be needed for diagnosis and treatment.

The incidence and histo-pathological characteristics of colorectal cancer in a population based cancer registry in Zimbabwe.

BACKGROUND: Data on colorectal cancer (CRC) in sub-Saharan Africa is mainly based on hospital series which suggest low incidence and frequent early onset cancers. This study characterises colorectal cancer in a population-based cancer registry in Zimbabwe. METHODS: Cases of CRC recorded by the Zimbabwe National Cancer Registry between 2003 and 2012 were analysed. Demographic and pathological characteristics were compared according to ethnicity and age. Trends in age standardised incidence rates (ASR) were determined. RESULTS: There were 886 and 216 cases of CRC among black Africans and Caucasians respectively, and 26% of the black Africans were younger than 40 years. Signet ring cell carcinomas were more common among black Africans compared to Caucasians (4% vs 1%, p=0.027). ASR increased by 1.9%/year and 3.9%/year among black African males and females respectively. CONCLUSION: CRC incidence is rising among black Africans and has unique demographic and pathological characteristics.

A comparison of Clostridium difficile diagnostic methods for identification of local strains in a South African centre.

Accurate diagnosis of Clostridium difficile infection is essential for disease management. A clinical and molecular analysis of C. difficile isolated from symptomatic patients at Groote Schuur Hospital, South Africa, was conducted to establish the most suitable clinical test for the diagnosis and characterization of locally prevalent strains. C. difficile was detected in stool samples using enzyme-based immunoassays (EIA) and nucleic acid amplification methods, and their performance was compared with that of C. difficile isolation using direct selective culture combined with specific PCR to detect the C. difficile tpi gene, toxin A and B genes and binary toxin genes. Toxigenic isolates were characterized further by ribotyping. Selective culture isolated 32 C. difficile strains from 145 patients (22 %). Of these, the most prevalent (50 %) were of ribotype 017 (toxin A- B+) while 15.6 % were ribotype 001 (toxin A+B+). No ribotype 027 strains or binary toxin genes (cdtA and cdtB) were detected. The test sensitivities and specificities, respectively, of four commercial clinical diagnostic methods were as follows: ImmunoCard Toxins A & B (40 % and 99.1 %), VIDAS C. difficile Toxin A & B (50 % and 99.1 %), GenoType CDiff (86.7 % and 88.3 %) and Xpert C. difficile (90 % and 97.3 %). Ribotype 001 and 017 strains had a 100 % detection rate by Xpert C. difficile, 100 % and 93.3 % by GenoType CDiff, 75 % and 53.3 % by ImmunoCard and 75 % and 60 % by VIDAS, respectively. The overall poor performance of EIA suggests that a change to PCR-based testing would assist diagnosis and ensure reliable detection of locally prevalent C. difficile 017 strains.

Self-expanding metal stent placement for oesophageal cancer without fluoroscopy is safe and effective.

BACKGROUND: Self-expanding metal stents (SEMS) are widely used to palliate patients with oesophageal cancer. Placement is usually done under endoscopic and fluoroscopic guidance. We have developed an exclusively endoscopic technique to deploy these stents. This article documents the technique and periprocedural experience. PATIENTS AND METHODS: All patients who had SEMS placement for oesophageal cancer at Grey's Hospital, Pietermaritzburg, South Africa, over a 5-year period (2007-2011) were reviewed. Stenting was performed without radiological guidance using the technique documented in this article. At endoscopy, the oesophageal lesion was identified, dilated over a guidewire if necessary, and a partially covered stent was passed over the wire and positioned and deployed under direct vision. Data were captured from completed procedure forms and included demographics, tumour length, the presence of fistulas, stent size and immediate complications. RESULTS: A total of 480 SEMS were inserted, involving 453 patients, of whom 43 required repeat stenting. There were 185 female patients (40.8%) and 268 male patients (59.2%). The mean age was 60 years (range 38 - 101). There were 432 black patients (95.4%), 15 white patients (3.3%) and 6 Indian patients (1.3%). The reasons for palliative stenting were distributed as follows: age>70 years n=95 patients, tumour>8 cm n=142, tracheo-oesophageal fistula (TOF) n=29, and unspecified n=170. One patient refused surgery, and one stent was placed for a post-oesophagectomy leak. Repeat stenting was for stent migration (n=15), tumour overgrowth (n=26) and a blocked stent and a stricture (n=1 each). Complications were recorded in six cases (1.3%): iatrogenic TOF (n=2), false tracts (n=3) and perforation (n=1). All six were nevertheless successfully stented. There was no periprocedural mortality. CONCLUSION: The endoscopic placement technique described is a viable and safe option with a low periprocedural complication rate. It is of particular use in situations of restricted access to fluoroscopic guidance.

Using a hybrid electronic medical record system for the surveillance of adverse surgical events and human error in a developing world surgical service.

INTRODUCTION: The quantification and analysis of adverse events is essential to benchmark surgical outcomes and establish a foundation for quality improvement interventions. We developed a hybrid electronic medical record (HEMR) system for the accurate collection and integration of data into a structured morbidity and mortality (M&M) meeting. METHODOLOGY: The HEMR system was implemented on January 1, 2013. It included a mechanism to capture and classify adverse events using the ICD-10 coding system. This was achieved by both prospective reporting by clients and by retrospective sentinel-event-trawling performed by administrators. RESULTS: From January 1, 2013 to March 20, 2014, 6,217 patients were admitted within the tertiary surgical service of Greys Hospital. A total of 1,314 (21.1 %) adverse events and 315 (5.1 %) deaths were recorded. The adverse events were divided into 875 "pathology-related" morbidities and 439 "error-related" morbidities. Pathology-related morbidities included 725 systemic complications and 150 operative complications. Error-related morbidities included 257 cognitive errors, 158 (2.5 %) iatrogenic injuries, and 24 (1.3 %) missed injuries. Error accounted for 439 (33 %) of the total number of adverse events. A total of 938 (71.4 %) adverse events were captured prospectively, whereas the remaining 376 (28.6 %) were captured retrospectively. The ICD-10 coding system was found to have some limitations in its classification of adverse events. CONCLUSIONS: The HEMR system has provided the necessary platform within our service to benchmark the incidence of adverse events. The use of the international ICD-10 coding system has identified some limitations in its ability to classify and categorise adverse events in surgery.

The role of endoscopic retrograde pancreatography in pancreatic trauma: a critical appraisal of 48 patients treated at a tertiary institution.

BACKGROUND: Endoscopic retrograde pancreatography (ERP) is useful in the diagnosis and treatment of selected patients with pancreatic trauma. We analyzed the role of ERP in treating persistent complications of pancreatic injuries at a tertiary institution. METHODS: Patients with pancreatic trauma who underwent ERP were identified from a prospective database of 426 pancreatic injuries from January 1983 to January 2011. Patient demographics, mechanism of injury, time to presentation, method of diagnosis, associated injuries, clinical management, endoscopic interventions and their timing, surgical treatment, and patient outcomes were evaluated. RESULTS: Forty-eight patients underwent ERP after blunt (n = 26) or penetrating (n = 22) pancreatic injury. Median time from injury to ERP was 38 days (range, 2-365 days). Diagnostic ERP was successful in 47 patients. In 11 patients, ERP demonstrated an intact main duct with minor peripheral injuries, and no further intervention was required. A pancreatic fistula was demonstrated in 24, a main pancreatic duct stricture in 12, and a pseudocyst in 10 patients. Fifteen patients had a pancreatic duct sphincterotomy, seven had a pancreatic stent inserted, and six had an endoscopic pseudocyst drainage. Ten patients ultimately required surgery, seven of whom had demonstrated a severe pancreatic duct stricture. Operations performed following ERP were distal pancreatectomy (n = 6), pancreaticojejunostomy (n = 3) and cyst-jejunostomy (n = 1). CONCLUSION: ERP allowed one quarter of the patients to be treated conservatively. Half had a successful intervention by ERP. Success was most likely in those with fistulae and pseudocysts. Surgery was ultimately avoided in more than three quarters of the patients. LEVEL OF EVIDENCE: Therapeutic study, level V.

The implications of the patterns of error associated with acute trauma care in rural hospitals in South Africa for quality improvement programs and trauma education.

INTRODUCTION: This audit uses error theory to analyze inappropriate trauma referrals from rural district hospitals in South Africa. The objective of the study is to inform the design of quality improvement programs and trauma educational programs. METHODS: At a weekly metropolitan morbidity and mortality meeting all trauma admissions to the Pietermaritzburg Metropolitan Trauma Service are reviewed. At the meeting problematic and inappropriate referrals and cases of error are identified. We used the (JCAHO) taxonomy to analyze these errors. RESULTS: During the period July 2009-2011 we received 1512 trauma referrals from our rural hospitals. Of these referrals we judged 116 (13%) to be problematic. This group sustained a total of 142 errors. This equates to 1.2 errors per patient. There were 87 males and 29 females in this group. The mechanism of injury was as follows, blunt trauma (66), stabs (32), gunshot wounds (GSW) (13) and miscellaneous five. The types of error consisted of assessment errors (85), resuscitation errors (26), logistics errors (14) and combination errors (17). The cause of the errors was planning failure in 68% of cases and execution failure in the remaining 32% of cases. The assessment errors involved the abdomen (50), chest (9), vascular system (8) and miscellaneous (18). The resuscitation errors involved airway (4), chest (11), vascular access (8) and cervical spine immobilization (3). CONCLUSIONS: Rural areas are error prone environments. Errors of execution revolve around the resuscitation process and current trauma courses specifically address these resuscitation deficits. However planning or assessment failure is the most common cause of error with blunt trauma being more prone to error of assessment than penetrating trauma.