SAMJ: South African Medical Journal

Collaboration is key to strengthening surgical research capacity in sub-Saharan Africa

Editor's choice February 2016

Difficult tracheal tube insertion: A new phraseology

Who will guard the guards? Medical leadership and conflict of interest in South African healthcare

The UCT class of 2000 reunion

Managing the fiscal beast - admin-to-clinician ratio now 3:1

Consciously cutting to the bone of SA's surgical/anaesthetic delivery

Surgery and anaesthesia in the South African context: Looking forward

Data mining and biological sample exportation from South Africa: A new wave of bioexploitation under the guise of clinical care?

Implications of direct-to-consumer whole-exome sequencing in South Africa

Challenging the cost of clinical negligence

Cardiovascular medicine in primary healthcare in sub-Saharan Africa: Minimum standards for practice (part 2)

An approach to a patient with infective endocarditis

Although infective endocarditis (IE) is relatively uncommon, it remains an important clinical entity with a high in-hospital and 1-year mortality. It is most commonly caused by viridans streptococci. Staphylococcus aureus is responsible for a malignant course of IE and often requires early surgery to eradicate. Other rarer causes are various bacilli, including the HACEK (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella and Kingella spp.) group of organisms and fungi. The clinical presentation varies. Patients may present with a nonspecific illness, valve dysfunction, heart failure (HF) and symptoms due to peripheral embolisation. The diagnosis is traditionally based on the modified Duke criteria and rests mainly on clinical features and to a lesser extent on certain laboratory findings, microbiological assessment and cardiovascular imaging. Identification of the offending micro-organism is not only important from a diagnostic point of view, but also makes targeted antibiotic treatment possible and provides useful prognostic information. A significant proportion of microbiological cultures are negative, frequently owing to the administration of antibiotics prior to appropriate culture. Blood-culture-negative IE poses significant diagnostic and treatment challenges. The course of the disease is frequently complicated, and sequelae include HF, local intracardiac extension of infection (abscess, fistula, pseudoaneurysm), stroke and intracranial haemorrhage due to septic emboli or mycotic aneurysm formation as well as renal injury. Management includes prolonged intravenous antibiotics and consideration for early surgery with removal of infective tissue and valve replacement in patients who have poor prognostic features or complications. Antibiotic administration for at-risk patients to prevent bacteraemia during specific procedures (particularly dental) is recommended to prevent IE. The patient population who would benefit from antibiotic prophylaxis has become increasingly restricted, and guidelines recommend prophylaxis only for patients with cyanotic congenital heart disease, prosthetic heart valves and a previous episode of IE. The management of a patient with IE is challenging and often requires multidisciplinary input from an IE heart team, which includes cardiologists.

An approach to the patient with suspected pericardial disease

Diseases of the pericardium commonly manifest in one of three ways: acute pericarditis, pericardial effusion and constrictive pericarditis. In the developed world, the most common cause of acute pericarditis is viral or idiopathic disease, while in the developing world tuberculous aetiology, particularly in sub-Saharan Africa, is commonplace owing to the high prevalence of HIV. This article provides an approach to the diagnosis, investigation and management of these patients.

An effective approach to chronic kidney disease in South Africa

Very few patients with end-stage kidney disease in South Africa receive renal replacement treatment (RRT), despite the rapidly growing demand, because of resource constraints. Nephrologists who agonise daily about who to treat and who not to, and have been doing so since the inception of dialysis in this country, welcomed the opportunity to interact with the National Department of Health at a recent summit of stakeholders. The major challenges were identified and recommendations for short- to long-term solutions were made. While the renal community can still improve efficiencies, it is clear that much of the responsibility for improving access to RRT and reducing inequities must be borne by the national government. The summit marks the first step in a process that we hope will ultimately culminate in universal access to RRT for all South Africans.

The decolonialisation of medicine in South Africa: Threat or opportunity?

The South African Traditional Health Practitioners Act 22 of 2007 is now fait accompli. The Act has been promulgated and the Department of Health (DoH) is proceeding with its implementation. An Interim Traditional Health Practitioners Council and a dedicated DoH deputy director have been appointed, the appointment of a registrar is being finalised, and the DoH has conducted a roadshow to introduce the Act and its implications to groups of traditional health practitioners (THPs) countrywide. The objective is eventual formalisation and professionalisation of THP practice to provide appropriate primary healthcare services through co-operation with biomedical service providers. Biomedical practitioners should understand the provisions of Act 22, and how this may affect their own practices.

Normal-pressure hydrocephalus presenting with psychiatric symptoms

A 62-year-old man presented with a 2-month history of psychiatric symptoms. These were preceded by cognitive deterioration, urinary incontinence and an abnormal gait. A diagnosis of normal-pressure hydrocephalus (NPH) was made, and the patient improved after surgery.

A multicentre evaluation of emergency abdominal surgery in South Africa: Results from the GlobalSurg-1 South Africa study

BACKGROUND: GlobalSurg-1 was a multicentre, international, prospective cohort study conducted to address the global lack of surgical outcomes data. Six South African (SA) hospitals participated in the landmark surgical outcomes study. In this subsequent study, we collated the data from these six local participants and hypothesised that the location of surgery was an independent risk factor for an adverse outcome following emergency intraperitoneal surgery. METHODS: Participating hospitals contributed 30-day outcomes data of consecutive emergency intraperitoneal surgical operations performed during a 2-week period between July and November 2014. The six heterogeneous hospital cohorts were compared by categorical confounders. The primary outcome measure was in-hospital mortality; secondary outcome measures were in-hospital morbidity and length of stay of >14 days. The unadjusted association between hospital and adverse outcome and the univariate association between categorical confounders and adverse outcome were tested. Significant associations were further tested by a multivariate stepwise forward logistic regression model built for each outcome of interest. RESULTS: Six hospitals (designated 1 - 6) contributed outcomes data for 169 operations. The mean age of the patients was 34.9 years (range 9 - 82), 116 (68.6%) were male, and the majority (37.2%) presented as a result of trauma. Hospital 5 was associated with 76-fold increased odds of in-hospital death and 58-fold increased odds of a major in-hospital complication, and hospital 3 was associated with 3-fold increased odds of any in-hospital complication. The final model predicting in-hospital death had a receiver operating characteristic curve statistic of 0.8892. CONCLUSION: The hospital is an independent risk factor for risk-adjusted adverse outcomes following emergency intraperitoneal surgeryin SA.

South African surgical registrar perceptions of the research project component of training: Hope for the future?

BACKGROUND: The Health Professions Council of South Africa requires that a research project be submitted and passed before registration as a specialist. OBJECTIVE: To describe surgical registrars' perceptions of the compulsory research project. METHOD: Ethics clearance was received before commencing the study. A questionnaire was developed to collect feedback from surgical registrars throughout South Africa (SA). Completed questionnaires underwent descriptive analysis using MS Excel. Fisher's exact test and the Χ² test were used to compare perceptions of the research-experienced and research-naive groups. RESULTS: All medical schools in SA were sampled, and 51.5% (124/241) of surgical registrars completed the questionnaire. Challenges facing registrars included insufficient time (109/124), inadequate training in the research process (40/124), inadequate supervision (31/124), inadequate financial resources (25/124) and lack of research continuity (11/124). Of the registrars sampled, 67.7% (84/124) believed research to be a valuable component of training. An overwhelming percentage (93.5%, 116/124) proposed a dedicated research block of time as a potential solution to overcoming the challenges encountered. Further proposals included attending a course in research methodology (79/124), supervision by a faculty member with an MMed or higher postgraduate degree (73/124), and greater research exposure as an undergraduate (56/124). No statistically significant differences were found between the perceptions of the research-experienced and research-naive groups. CONCLUSIONS: Challenges facing surgical registrars in their efforts to complete their research projects were identified and solutions to these problems proposed. It is heartening that respondents have suggested solutions to the problems they encounter, and view research as an important component of their careers.

Favourable outcomes for the first 10 years of kidney and pancreas transplantation at Wits Donald Gordon Medical Centre, Johannesburg, South Africa

BACKGROUND: It is important for centres participating in transplantation in South Africa (SA) to audit their outcomes. Wits Donald Gordon Medical Centre (WDGMC), Johannesburg, SA, opened a transplant unit in 2004. The first 10 years of kidney and pancreas transplantation were reviewed to determine outcomes in respect of recipient and graft survival. METHODS: A retrospective review was conducted of all kidney-alone and simultaneous kidney-pancreas (SKP) transplants performed at WDGMC from 1 January 2004 to 31 December 2013, with follow-up to 31 December 2014 to ensure at least 1 year of survival data. Information was accessed using the transplant registers and clinical records in the transplant clinic at WDGMC. The Kaplan-Meier method was used to estimate 1-, 5- and 10-year recipient and graft survival rates for primary (first graft) kidney-alone and SKP transplants. RESULTS: The overall 10-year recipient and graft survival rates were 80.4% and 66.8%, respectively, for kidney-alone transplantation. In the kidney-alone group, children tended towards better recipient and graft survival compared with adults, but this was not statistically significant. In adults, recipient survival was significantly better for living than deceased donor type. Recipient and graft survival were significantly lower in black Africans than in the white (largest proportion in the sample) reference group. For SKP transplants, the 10-year recipient survival rate was 84.7%, while kidney and pancreas graft survival rates were 73.1% and 43.2%, respectively. CONCLUSION: Outcomes of the first 10 years of kidney and pancreas transplantation at WDGMC compare favourably with local and international survival data.

Factors determining clinical outcomes in intussusception in the developing world: Experience from Johannesburg, South Africa

BACKGROUND: Rates of open reduction of intussusception were noted to be unacceptably high during an institutional internal audit OBJECTIVES: To determine the impact of revised protocols to better select patients for pneumatic reduction (PR), and document associated morbidity and mortality, and the factors that affect the above METHODS: Medical records of patients between 3 months and 3 years of age presenting to the Department of Paediatric Surgery at Chris Hani Baragwanath Academic Hospital, Johannesburg, South Africa, from 2007 to 2010 were reviewed. Determining factors, including duration of symptoms, admission C-reactive protein (CRP) level and weight, were analysed against clinical outcomes, notably PR, bowel resection, relook laparotomy and death RESULTS: A total of 97 cases were suitable for inclusion. In 62 of these (63.9%), PR was attempted; this was successful in 32 cases (51.6%), giving an overall successful PR rate of 33.0%. In 7 of the 62 patients, a pneumoperitoneum was documented during the reduction attempt. Of the 65 patients who underwent surgery, 53 required intestinal resection and 12 had spontaneous or manual reduction. Ileostomy was necessary in 9 patients, and 7 required relook laparotomy. The overall mortality rate was 9.1%. Averages of 'determining factors' assessed against clinical outcome were as follows: mean weight (standard deviation (SD)) 7.4 (4.3) kg, mean duration of symptoms (DOS) 3.0 (SD 2.2) days, and admission CRP level 50.9 mg/L (range 1 - 249.3). Prolonged DOS and a raised CRP level predicted a poor outcome CONCLUSIONS: Despite marked improvements in management and PR outcomes, intussusception remains associated with significant morbidity and mortality. Prolonged DOS and an elevated CRP predict worse outcomes. The use of these markers in association with clinical factors may assist management decisions, specifically with regard to operative or non-operative management. Awareness and education are key to prompt presentation and early diagnosis. Well-defined protocols introduced at all points of contact ensure early recognition and resuscitation as well as prompt referral for definitive management

Are central hospitals ready for National Health Insurance? ICD coding quality from an electronic patient discharge record for clinicians

BACKGROUND: South Africa (SA)'s planned National Health Insurance reforms require the use of International Statistical Classification of Diseases (ICD) codes for hospitals to purchase services from the proposed National Health Authority. However, compliance with coding at public hospitals in the Western Cape Province has been challenging. A computer application was developed to aid clinicians in integrating ICD coding into the patient hospital discharge process. OBJECTIVES: To evaluate the quality of ICD codes captured using the application and predictors thereof in a single hospital department. METHODS: After 6 months, the quality of ICD codes was determined by comparing ICD code descriptors with medical concepts in a random sample of original patient records selected over a 6-week period. Patient and personnel characteristics influencing quality of coding, derived from a theoretical framework, were collected. RESULTS: Of 223 patient records, 45.3% (95% confidence interval (CI) 38.8 - 51.9) had complete ICD codes. Primary ICD code accuracy was 74.0% (95% CI 67.8 - 79.5). Patient characteristics such as female gender, younger age group and fewer comorbidities, as well as seniority of clinician rank, were significantly associated with ICD coding being complete on adjusted analysis. CONCLUSIONS: The results of this study describe ICD coding quality at a central hospital in SA supported by a computer application and the factors influencing this. More interventions are required to achieve reliable coding data, such as additional ICD coding validation tools, training and oversight of junior clinicians.

Burn surgeons in South Africa: A rare species

BACKGROUND: The high burden of burn injuries in South Africa (SA) requires surgeons skilled in burn care. However, there are few dedicated burn surgeons and properly equipped units or centres. OBJECTIVES: To quantify the involvement of surgeons in burn care in SA hospitals, identify factors that attract surgeons to pursue burn care as a career and deter them from doing so, and understand the challenges of hospitals treating burn patients around the country. METHODS: This was a prospective, qualitative study. Questionnaires were handed out at the South African Burn Society Congress in September 2013 and a trade symposium in March 2014. RESULTS: One hundred questionnaires were handed out, and there was a 70% response rate. Twenty-six (39%) of the respondents had a specialist surgical qualification. Only half the units had registrars (48%) and interns (51%) on their staff. Only 30% of the respondents were dedicated to burn care alone, the majority being involved on a part-time basis. The most common factor respondents suggested was needed to recruit future burn care providers, cited by 76%, was better facilities and resources. Other factors included training and skills development (59%), subspecialist training (55%), development of a diploma in burn care (52%), development of research (52%) and healthcare worker psychological support (45%). CONCLUSION: We have demonstrated that current workforce resources for burn care are inadequate, the major deficit being lack of training and the resource-restricted environment. This survey provides basic information towards workforce planning, which can be used to inform the necessary strategic decisions.

Mortality in paediatric burns victims: A retrospective review from 2009 to 2012 in a single centre

BACKGROUND: Childhood mortality is high in low- and middle-income countries. Burns are one of the five leading causes of childhood injury mortality in South Africa (SA). While there is an abundance of literature on burns in the developed world, there are far fewer publications dealing with childhood mortality related to burns in Africa and SA. OBJECTIVE: To describe the mortality of children admitted to a dedicated paediatric burns unit, and investigate factors contributing to reducing mortality. METHODS: A retrospective review was performed of patients admitted to the Johnson & Johnson Paediatric Burns Unit, Chris Hani Baragwanath Academic Hospital, Johannesburg, SA, between May 2009 and April 2012. RESULTS: During the study period, 1 372 patients aged <10 years were admitted to the unit. There were 1 089 admissions to the general ward and 283 admissions to the paediatric burns intensive care unit (PBICU). The overall mortality rate was 7.9% and the rate for children admitted to the PBICU 29.3%; 90.8% of deaths occurred in children aged <5 years. Of children admitted with an inhalational injury, 89.5% died. No child with a burn injury >60% of total body surface area (TBSA) survived. CONCLUSIONS: Our overall mortality rate was 7.9%, and the rate declined significantly over the 3-year study period from 11.7% to 5.1%. Age <5 years, the presence of inhalational injury, burn injury >30% of TBSA and admission to the PBICU were significant risk factors for mortality.

Validating homicide rates in the Western Cape Province, South Africa: Findings from the 2009 Injury Mortality Survey

BACKGROUND: The Western Cape Province had the highest homicide rates in South Africa during the early 2000s. South African Police Service (SAPS) data suggested a significant decline in homicide rates in the Western Cape since 2007. It ranked second highest to the Eastern Cape Province until 2013 and ranked highest again at 52.1/100 000 in 2015. A recent national injury mortality survey offers an alternative data source to assess whether the decline in homicide rates in the Western Cape was real. METHODS: A retrospective record review of autopsies was conducted from 45 state mortuaries in eight provinces for 2009. In addition, mortality data for the Western Cape were sourced from the Provincial Injury Mortality Surveillance System. Age-standardised mortality rates and crude homicide rates per 100 000 population were calculated to compare with the SAPS crude rates. RESULTS: Our study found that the Western Cape had a provincial age-standardised homicide rate of 40.1/100 000 in 2009 and ranked fourth highest among the nine provinces. The crude homicide rate of 43/100 000 for the Western Cape was similar to the SAPS provincial homicide rate of 42.4/100 000. The Northern Cape Province was the only notable exception to our provincial homicide rate ranking comparison with the SAPS for 2009. CONCLUSIONS: The Western Cape is fortunate to have alternative data sources to monitor trends in homicides over time. The latest release of the 2014/2015 SAPS crime statistics should be assessed in a similar manner, with a more recent data source, to validate accuracy of the provincial rates on a regular basis.

Empirical antimicrobial therapy for probable v. directed therapy for possible ventilator-associated pneumonia in critically injured patients

BACKGROUND: Ventilator-associated pneumonia (VAP) has recently been classified as possible or probable. Although direct attributable mortality has been difficult to prove, delay in instituting appropriate therapy has been reported to increase morbidity and mortality. Recent literature suggests that in possible VAP, instituting directed therapy while awaiting microbiological culture does not prejudice outcome compared with best-guess empirical therapy. OBJECTIVES: To ascertain outcomes of directed v. empirical therapy in possible and probable VAP, respectively. METHODS: Endotracheal aspirates were obtained from patients with suspected VAP. Those considered to have possible VAP were given directed therapy following culture results, whereas patients with more convincing evidence of VAP were classed as having probable VAP and commenced on empirical antimicrobials based on microbiological surveillance. RESULTS: Pneumonia was suspected in 106 (36.8%) of 288 patients admitted during January - December 2014. Of these, 13 did not fulfil the criteria for VAP. Of the remaining 93 (32.2%), 31 (33.3%) were considered to have probable and 62 (66.7%) possible VAP. The former were commenced on empirical antimicrobials, with 28 (90.3%) receiving appropriate therapy. Of those with possible VAP, 34 (54.8%) were given directed therapy and in 28 (45.2%) no antimicrobials were prescribed. Of the latter, 24 recovered without antimicrobials and 4 died, 3 from severe traumatic brain injury and 1 due to overwhelming intra-abdominal sepsis. No death was directly attributable to failure to treat VAP. No significant difference in mortality was found between the 34 patients with possible VAP who were commenced on directed therapy and the 31 with probable VAP who were commenced on empirical antimicrobials (p=0.75). CONCLUSIONS: Delaying antimicrobial therapy for VAP where clinical doubt exists does not adversely affect outcome. Furthermore, this policy limits the use of antimicrobials in patients with possible VAP following improvement in their clinical condition despite no therapy.

An analysis of patients transported by a private helicopter emergency medical service in South Africa

BACKGROUND: A helicopter emergency medical service (HEMS) is a specialist flying emergency service where on-board medical personnel have both the knowledge and equipment to perform complicated medical procedures. The paucity of literature describing the types of patients flown by HEMS in South Africa (SA) and their clinical outcome poses a challenge for current aeromedical services, as there is no baseline information on which to base flight criteria, staffing and policy documents. This has the potential to hamper the advancement of HEMS in SA. OBJECTIVES: To undertake a descriptive analysis of patients flown by the Netcare 911 HEMS over a 12-month period in Gauteng and KwaZulu-Natal (KZN) provinces, SA, and to assess patient outcomes. The clinical demographics of patients transported by the HEMS were analysed, time frames from dispatch of the helicopter to delivery of the patient to the receiving hospital determined, and patient outcomes at 24 hours and 72 hours analysed. METHODS: The study utilised a retrospective quantitative, descriptive design to analyse patients transported by a private HEMS in SA. All complete records of patients transported by the Netcare 911 HEMS between 1 January and 31 December 2011 were included. RESULTS: The final study population comprised 537 cases, as 10 cases had to be excluded owing to incomplete documentation. Of the 537 cases, 82 (15.3%) were managed by the KZN HEMS and 455 (84.7%) by the Gauteng HEMS. Adult males were the patients most commonly flown in both Gauteng and KZN (350/455 patients (76.9%) in Gauteng and 48/82 (58.5%) in KZN were males, and 364/455 patients (80.0%) in Gauteng and 73/82 (89.0%) in KZN were adults). Motor vehicle collisions were the most common incidents necessitating transport by HEMS in both operations (n=193, 35.9%). At the 24-hour follow-up, 339 patients (63.1%) were alive and stable, and at the 72-hour follow-up, 404 (75.3%) were alive and stable. CONCLUSIONS: The study findings provided valuable information that may have an impact on the current staffing and authorisation criteria of SA HEMS operations.

Treatment and outcome of unusual animal bite injuries in young children

BACKGROUND: Animal bites are a major cause of preventable traumatic injuries. OBJECTIVES: To provide more epidemiological information on animal bites, and assist in increasing awareness of the problem. METHODS: A retrospective chart review was performed including children aged >13 years presenting with bite injuries (excluding dog and human bites) to the trauma unit at Red Cross War Memorial Children's Hospital, Cape Town, South Africa, over a 25-year period. RESULTS: Two hundred and thirteen children were eligible to be entered into the study. The median age was 2.9 years (range 1.2 - 6.5), with boys slightly predominating (54.9%). Most (74.6%) of the bite injuries were inflicted by mammals, the majority (64.8) of mammalian bites being rat bites. The proportions of boys and girls in the age group 0 - 4 years bitten by rats significantly differed from the proportions in the age group >4 years (p=0.039). In the age group 0 - 4 years more girls suffered rat bites, while more boys were bitten in the age group >4 years. Of 91 rat bites, 81 (89.0%) occurred inside the house. The hands (43.9%) and the head/face/neck region (39.0%) were most affected. The underdeveloped suburbs of Philippi, Gugulethu and Khayelitsha in Cape Town represented a disproportionate number (41.6%) of rat bites. CONCLUSIONS: There is a relationship between poverty, unemployment, poor housing, informal settlements and rodent infestation. These high-risk populations need to be the target for government rat eradication programmes.

Thrombocytopenia during pregnancy in women with HIV infection receiving no treatment

BACKGROUND: Thrombocytopenia (TCP) complicates 5 - 8% of pregnancies. Most cases of TCP are gestational, and the condition is usually mild and occurs in the latter part of pregnancy. Apart from pregnancy-associated medical complications such as pre-eclampsia, HIV infection is a recognised cause of TCP, and a relatively high prevalence of TCP during pregnancy would be expected in a setting with a high antenatal seroprevalence of HIV. METHODS: This was a sub-analysis of the data from a prospective trial in which the incidence of postpartum sepsis in HIV-infected women was compared with that in HIV-uninfected women. Women who were considered at low risk and eligible for vaginal delivery were recruited at 36 weeks' gestation, and followed up for 6 weeks after delivery. Full blood counts and CD4 counts of HIV-infected women were obtained at baseline and repeated 6 weeks after delivery. RESULTS: The prevalence of TCP was 5.3% during pregnancy and 1.2% 6 weeks after delivery. The prevalence was similar among HIV-infected (6.0%) and HIV-uninfected women (4.7%) (p=0.292). Among the HIV-infected women, who were not receiving antiretroviral therapy (mean CD4 cell count of 453 cells/μL), there was no significant association between immunosuppression and the severity of TCP. CONCLUSIONS: Most of the TCP seen during pregnancy is of the gestational variety, and in this study HIV infection did not increase its prevalence or its severity.

Which test is best for diagnosing peanut allergy in South African children with atopic dermatitis?

BACKGROUND: Diagnosing peanut allergy based on sensitisation alone leads to an unacceptable rate of overdiagnosis. OBJECTIVE: To define parameters that may help differentiate peanut allergy from asymptomatic sensitisation in a cohort of South African (SA) children with atopic dermatitis (AD). It is the first study in SA to utilise oral food challenge tests and analyse peanut component patterns. METHODS: This was a prospective, observational study at a paediatric university hospital in Cape Town, SA. Children with AD, aged 6 months - 10 years, were recruited randomly. They were assessed for sensitisation and allergy to peanut by questionnaire, skin-prick tests (SPTs), immuno solid-phase allergen chip (ISAC) tests, ImmunoCAP component tests to Ara h 1, 2, 3, 8 and 9, and incremental food challenges. RESULTS: One hundred participants (59 Xhosa (black Africans) and 41 of mixed race, median age 42 months) were enrolled. Overall, 44% of patients were peanut sensitised and 25% had a true peanut allergy. SPTs and ImmunoCAP Ara h 2 produced the highest areas under the receiver operating characteristic curve for predicting peanut allergy in peanut-sensitised patients. The ISAC test was less sensitive, more specific and produced significantly lower median values than ImmunoCAP tests. Ara h 2 was the most useful component in differentiating allergy from tolerance in both ethnic groups, being positive in 92% of allergic and 40% of sensitised but tolerant children (p<0.001). There was little additional contribution from Ara h 1 and 3. Ara h 8 and 9 were associated with tolerance. Commonly used 95% positive predictive values (PPVs) for SPTs, peanut-specific IgE and Ara h 2 levels fared suboptimally in our population. Maximum PPVs for this study population were found at SPT 11 mm, peanut IgE 15 kU/L and ImmunoCAP Ara h 2 of 8 kU/L, but these adjusted levels still had suboptimal PPVs in Xhosa subjects. Severe peanut allergy was associated with increased median peanut IgE and Ara h 2. CONCLUSIONS: The component Ara h 2 was useful for differentiating allergy from tolerance in both ethnic groups in this SA cohort. Ninety-five percent PPVs for peanut allergy tests may need to be revised, especially in Xhosa patients. An SPT result >11 mm as well as Ara h 2 >8 kU/L had the best predictive value for peanut allergy.