Table 1

Results of the analysis, arranged by type of study (prospective or retrospective) and year of publication

Author, year, referenceCountryDesignMethodsFindingsMeasure type
Prospective studies
Mantel et al, 199814South AfricaDescriptiveDaily chart audit of 13 429 deliveries for maternal near misses and mortality over 1 yearNear miss was defined as a patient with acute organ system dysfunction that, if not treated, could lead to death. Of 13 429 deliveries, the study found 147 maternal near misses and 30 maternal deaths. Review processes incorporated into normal chart review identified opportunities for improvement.
Mbanya et al, 200110CameroonDescriptiveBlood bank and patient record review and observation for transfusion related adverse events involving 40 134 units of blood transfused over 5 years50% of transfusions were associated with an “unfavorable outcome” (febrile reaction, urticaria, acute haemolytic reactions, delayed haemolytic reactions, circulatory overload, infections or death). Only basic screening mechanisms were used to screen for disease or compatibility.
Nogueira et al, 200215BrazilDescriptiveAudit of 164 cardiac patients for postsurgical hepatitis C infection for 6 monthsThe incidence of postsurgical hepatitis C infection was 1.27%.
Hatherill et al, 200313South AfricaProspective CohortAudit of 202 consecutive paediatric transports for adverse events over 1 yearOne or more technical, clinical or critical adverse events occurred in 36%, 27% and 9% of transports, respectively. Technical events were not associated with mortality. Transfers from non-academic metropolitan hospitals had significantly higher incidence of adverse events.
English et al, 200416KenyaDescriptiveAudit of 14 government hospitals for performance measuresEvidence of high mortality rates, inaccurate prescribing and diagnosing practices and inappropriate care provision was found. Incomplete records, insufficient training/diagnosis and medication errors contributed to high mortality rates and low-quality care.
Krug et al, 200417South AfricaDescriptiveReview of the causes of deaths that occurred over 1 year in patients <5 years old74% of deaths were in children aged 1 to 11 months. The most common causes of death were lower respiratory tract infection, AIDS, sepsis and malnutrition. Contributing administrative factors were a missing “road-to-health card”, communication problems and insufficient documentation.
Vong et al, 20059CambodiaDescriptiveSurvey of 500 patients and 60 providers and direct observation examining injection practicesOverall injection rate was 5.9 injections per person per year. Prescribers reported that 47% of prescribed medications required injection. 13% of observed providers did not dispose of the needle or syringe appropriately.
dos Santos and Coelho, 200612BrazilProspective cohortContinuous audit, medical record review and family member survey examining ADRs in 272 paediatric patients over 5 months12.5% of children experienced at least one ADR during a hospital stay. Patient factors associated with risk of ADR were number of drugs administered and previous hospitalisations. 78.6% of ADRs were classified as moderate to severe.
Kermode and Muani, 20067IndiaDescriptiveSurvey of 280 patients and 120 family members examining injection practices over 15 months47.1% of inpatients and 81.5% of their relatives reported injection from an unqualified provider or relative. Most reported safe injection practices. 7.6% reported that needles or syringes were not sterile.
Rosenthal et al, 2006188 developing countriesProspective cohortAudit of 55 ICUs for device-associated infection rates over 3 yearsThe overall rate of infection was 14.7%. Ventilator-associated pneumonia was most common (41% of infections). Crude mortality of patients with device-associated infection was 44.9% for ventilator-associated pneumonia.
Chamisa and Zulu, 20078South AfricaDescriptiveReview of 204 randomly selected case notes in a 2-month period examining adherence to the Royal College of Surgeons of England case note guidelines16/35 guidelines were met with 80% compliance. Non-compliance with standards included lack of patient, clinician and hospital identifying information and date as well as lack of clinical information such as complaint, drug history, family history and allergies.
Pandey et al, 200711IndiaDescriptiveAudit of phlebotomy errors with 11 260 donors over 8 monthsThe authors found a 3.1% incidence of errors. 55.1% of errors were technical and 44.9% were clerical. 42% of technical errors lead to donor injury or discarding of collected unit. Trained staff were responsible for all major errors.
Retrospective studies
Kambarami et al, 200023ZimbabweRetrospective cohortCase note review of 234 patients admitted during 1 year for neonatal mortalityCase fatality was 46.4%. Odds of death for infants with respiratory distress treated with ventilation were 12.29 times those treated with CPAP.
Sharma et al 200128IndiaDescriptiveReview of reported transfusion incidents for errors over 1 year123 errors were found. 87% occurred outside of the blood bank (bedside or in-transit); most were labelling errors. Problems with patient identification numbers resulted in serious morbidity for two patients. Technical error led to serious morbidity for one patient.
Pai, 200325IndiaDescriptiveSurvey of 50 pathologists and 47 clinicians for surgical pathology errors in last yearClinicians remembered twice the number of errors as pathologists. Pathologists remembered mostly lymphoid disease errors, whereas clinicians remembered mostly gastrointestinal tract and lymphoid errors. The authors suggested a need for better communication and data collection.
Carreazo et al 200419PeruDescriptiveReview of medical records of 167 of patients admitted in 1 month for primary diagnosis and treatment94% of interventions were evidence-based, but 75% of interventions were based on expert opinion or basic science.
Matsuoka et al, 200421BrazilDescriptiveMedical record review of eight Salmonella cases for patterns and opportunities for risk reduction over 6 weeksAll patients developed diarrhoea and received enteral nutrition. Outbreak control procedures were appropriate and cultures from nutrition employees were negative.
Kakkar et al, 200526IndiaDescriptiveCase record review of 82 patients for adequacy of anticoagulation processes and outcomesOutpatient anticoagulant control was inadequate, patients were undercoagulated (only 17.8% patients were in the therapeutic range) and adherence to guidelines was low. 19.5% of patients had treatment complications.
Hesse et al, 200529GhanaDescriptiveAdmission record, discharge record and case note review for clinical management and mortality in 158 patients with tetanus over 8 yearsThe overall fatality rate was 50%. Case fatality rate was highest in patients without wound debridement (76.2%), antibiotics (76.2%) and TIG (73.1%) and lowest in patients with TIG and antibiotics (14.3%).
Yan et al, 200622ChinaDescriptiveReview of 1452 medical records, prescription record review, direct observation and survey for injection practices over a 4-month periodThe unsafe injection rate was 16%, and the unnecessary injection rate was 56%. 96% of all needles and syringes were disposed of properly. 55% of injection providers reported injury by waste needles in the last year.
Mukhtar-Yola and Iliyasu, 200724NigeriaDescriptiveMedical record review of neonatal morbidity and mortality in 3014 admissions over 6 years16.9% of babies died. Leading diagnoses were birth asphyxia, neonatal sepsis and prematurity. Odds of death for babies delivered outside the hospital were 71% higher than those delivered inside the hospital. Neonatal morbidity and mortality were found to be preventable with antenatal care, supervised delivery, appropriate care and early referral.
Jimmy et al, 200720IndiaDescriptive3.5-year review of 2312 Drug Information Center queries for detail, purpose and category of ADRs25.9% of queries were related to adverse drug reactions. Most ADR queries pertained to identification of ADRs and uncommon reactions.
Khan and Khan, 200727PakistanDescriptiveMorbidity and mortality record review examining anaesthesia-related death over 12 yearsThe crude mortality rate was 3.14/10 000, anaesthesia-caused mortality rate was 0.35/10 000 and anaesthesia-related mortality was 0.7/10 000. Most deaths occurred in the postoperative period and in general surgery. Other contributing factors included inappropriate decision-making, inappropriate dose of narcotic, inadequate preoperative preparation and postoperative monitoring.
  • ADR, adverse drug reaction; CPAP, continuous positive airway pressure; ICU, intensive care unit; O, outcome; P, process; S, structure; TIG, tetanus immunoglobulin.