Time intervals in the treatment of fractured femurs as indicators of the quality of trauma systems
Amir Matityahu a, Iain Elliott b, Meir Marmor c, Amber Caldwell d, Richard Coughlin b & Richard A Gosselin e
a. San Francisco General Hospital, University of California in San Francisco, 2550 Twenty-third Street, Building 9, 2nd Floor, San Francisco, CA 94110, United States of America (USA).
b. University of Florida, Gainesville, USA.
c. Orthopaedic Trauma Institute, University of California, San Francisco, USA.
d. Institute for Global Orthopaedics and Traumatology, University of California, San Francisco, USA.
e. School of Public Health, University of California, Berkeley, USA.
Correspondence to Amir Matityahu (e-mail: firstname.lastname@example.org).
(Submitted: 07 March 2013 – Revised version received: 08 August 2013 – Accepted: 23 August 2013 – Published online: 04 October 2013.)
Bulletin of the World Health Organization 2014;92:40-50. doi: http://dx.doi.org/10.2471/BLT.13.120436
According to the global burden of disease study of 2010, injury accounted for 10% of deaths worldwide and 11.2% of all disability-adjusted life years (DALYs).1,2 Rising awareness among policy-makers of the burden of injury3 has led to an increased need for easily quantifiable metrics to improve the allocation of resources for the treatment of injuries.4–11 The intervals involved in the treatment of femur fractures – injury to admission, admission to surgery and surgery to discharge – are easily measurable and may be useful in evaluating the efficiency with which a trauma system treats an injured patient.
Since the poorest third of the world’s population accounts for only 3.5% of all of the surgical procedures performed, there is clearly a large disparity in access to surgical care.12,13 The annual incidence of injury-related deaths is also substantially higher in low- and middle-income countries than in high-income countries: 89 versus 51 per 100 000 inhabitants, respectively.3 Injuries caused by motorized vehicles are currently the tenth leading cause of DALYs, are projected to be the third highest cause of DALYs by 202014–19 and are the leading cause of death among individuals aged 15–29 years.1 In low- and middle-income countries, the rapid increase in industrialization – without concomitant improvements in infrastructure – is likely to lead to a rise in the incidence of injuries caused by motorized vehicles.
Resource allocation for the care of skeletal traumas differs both between and within countries.20 The consequences of these differences are difficult to quantify because there is a general lack of access to the many parameters that would be needed to make a detailed assessment of patient care within trauma systems. New metrics to evaluate the efficiency and efficacy of trauma systems are required as aids in decisions on resource allocation.4–9 A good indicator of the quality of care in a clinical environment must have relevancy, validity, reliability and feasibility.21 The relevancy of an indicator depends on how well it measures the problems that are experienced in the provision of care. Validity is demonstrated via strong correlations with other measures of the current quality of care. Reliability and feasibility refer to the variation in measurements – both inter- and intra-observer – and the ease with which the indicator can be measured, respectively.
Femur fractures are typically high-velocity injuries that are associated with high rates of both short- and long-term disability (Table 1).2,22,23 Unlike other fractures of the upper and lower extremities, which could occasionally be treated acceptably with splinting in non-hospital centres, operative management or traction is required for a good functional outcome after a femur fracture.22–24 Skeletal traction, which typically involves having the patient lie in bed for up to 3 months, prevents most patients from working and places a large burden on the patient’s family. In contrast, a patient may be ambulatory within a week of the operative management of a fractured femur by intramedullary nailing. Compared with traction, such nailing – which is now the standard treatment in high-income countries – allows a much faster return to normal functioning.23–25 The intramedullary nailing of femur fractures has been shown to be cost-effective in low- and middle-income countries and results in better patient outcomes than those seen with skeletal traction.24 Unlike more lethal forms of trauma, most isolated femur fractures are seen and recorded by some form of care provider.
In the overall management of femur fractures, the time from injury to hospital discharge can be segmented into three main treatment intervals: injury to admission, admission to surgery and surgery to discharge. The initial injury to admission interval is representative of the presence and efficacy of pre-hospital emergency medical systems, such as ambulance services. Pre-hospital time has been reported as a valid quality indicator in other research articles but not specifically for patients with fractured femurs.5,26 The admission to surgery interval is affected by in-hospital variables such as the availability of human resources and essential equipment and the hospital infrastructure.27 The final surgery to discharge interval may be affected by medication availability, nursing care, rehabilitation and the post-operative availability of social services.28,29
This study proposes that the intervals from injury to admission, admission to surgery and/or surgery to discharge in the care of traumatic, isolated fractures of the femoral shaft can be good indicators of the general quality of musculoskeletal trauma services and can be validated as indicators against national economic and health system parameters, which are the best available data for such validation.
We reviewed the medical records of patients with isolated femur fractures who had received operative treatment in six tertiary trauma hospitals outside the United States of America. The hospitals were the Medicine Horschule Hannover (Hannover, Germany), the Komfo Anokye Teaching Hospital (Kumasi, Ghana), the Hadassah Medical Center (Jerusalem, Israel), the Moi Teaching and Referral Hospital (Eldoret, Kenya), the Bedford Orthopaedic Hospital (Mthata, South Africa) and the Muhumbili Orthopaedic Institute (Dar es Salaam, United Republic of Tanzania). These hospitals were chosen because each of them had a long-standing partnership with the University of California’s Institute for Global Orthopaedics and Traumatology. The years investigated at each hospital depended on the records available.
Corresponding data for the United States – for the years 2002–2006 – were gathered from the United States National Trauma Data Bank.30 Only data from patients with isolated closed femur fractures, an injury severity score of < 10 and an age of more than 18 years were included. An injury severity score is calculated as the sum of the squares of the three highest measurements on the abbreviated injury scale. We used an injury severity score of < 10 as an inclusion criterion because – as an isolated femur fracture is assigned a value of 3 on the abbreviated injury scale – any patient with an isolated femur fracture and no additional injury would be given an injury severity score of 9. Any additional injury would increase this score to at least 10.31
Each country providing data was categorized as low-, middle- or high-income according to the criteria of the World Bank.32
Data gathering from the hospitals
Information was gathered from patient charts, patient interviews or the hospital trauma registry – when available. The study protocol was approved by the institutional review boards at the University of California in San Francisco and at the study hospitals with such boards.
Data variables collected for each patient
The data collected for each patient were age, sex, mechanism of injury, type of pre-hospital transportation, type of initial management, duration of initial management, type of definitive management, whether surgery had been open or percutaneous, length of surgery, type of anaesthesia, implants used, implant source, surgical complications, mean time to the performance of the definitive surgical intervention and the time intervals between injury and hospital admission, admission and surgery and surgery and discharge.
Hospital, resource and country data
For each study hospital, the hospital administrative teams helped enumerate inpatient and emergency department beds, hospital beds dedicated to orthopaedic patients, hospital admissions per year, orthopaedic admissions per year, functioning operating rooms, fluoroscopy machines in use, functioning computed tomography and magnetic resonance imaging scanners available for use, orthopaedic procedures per year, qualified orthopaedic surgeons, qualified general surgeons, qualified anaesthesiologists, orthopaedic officers and physical therapists. The extent of each hospital’s orthopaedic coverage – that is, whether such coverage was available 24 hours per day – was also recorded.
Student’s t-tests were used to compare the mean intervals from injury to admission, admission to surgery and surgery to discharge that were recorded in low- and middle-income countries with those that were recorded in high-income countries. In these tests, the means for each study site were weighted according to the number of study patients from the site.
Attempts were made to validate the time interval data – as indicators of the quality of the trauma system – by measuring how well the intervals correlated with various development parameters of the country involved. The parameters investigated were: population; adult literacy; life expectancy; mortalities among infants and children aged less than 5 years; the modelled number of road traffic deaths; estimated mortality from road traffic accidents; per capita values for total and general government expenditures on health; per capita and national values for the gross domestic product (GDP); the human development index; the Gini coefficient; the corruption index; out-of-pocket expenditure and money spent on private prepaid plans – as percentages of private expenditure on health; general government and private expenditures on health – as percentages of the total expenditure on health; and social security expenditure on health – as a percentage of general government expenditure on health.
Levels of correlation were evaluated by using the Excel software package (Microsoft, Redmond, USA) to calculate Pearson’s product moment correlation coefficients.
Overall, 4967 femur fractures – 4644 from the United States and 323 from other countries – met the inclusion criteria (Table 2). The included patients from the United States had a mean age of 38 years and 59% of them were male. Most (57%) of the included patients were the victims of road traffic accidents. In terms of mechanism of injury, mean age, use of intramedullary nailing and type of anaesthesia, the patients from high-income countries – Germany, Israel and the United States – were similar to those from low- and middle-income countries – Ghana, Kenya, South Africa and the United Republic of Tanzania. However, compared with a patient from a high-income country, a patient from a low- or middle-income country was significantly more likely to be male (P = 0.019) and to have undergone open nailing – rather than a percutaneous reduction (P = 0.001). The demographics, catchment areas and resource availability of the six study hospitals are summarized in Table 3. The internal resources of each hospital included in the study – taken as indicators of preparedness for surgery – are summarized in Table 4.
Table 2. Patient characteristics, treatment and transport in the study hospitals
Table 3. Demographics, catchment areas and funding sources pertaining to the study hospitals
Table 4. Internal resources of the study hospitals, indicative of preparedness for surgery
The intervals from injury to admission, admission to surgery and surgery to discharge were all longer in the low- and middle-income countries than in the high-income countries (Table 5). The relationships between each of the three types of intervals investigated and the corresponding country development parameters are summarized in Table 6.
Table 5. Time intervals involved in the treatment of femur fractures, by hospital/databank
Table 6. Coefficients for the correlations between time intervals involved in the treatment of femur fractures and country development indicators
There were strong inverse correlations (P < 0.05 for each) between the interval from injury to admission recorded in a country and that country’s general government expenditure on health – as a percentage of total expenditure on health; social security expenditure on health – as a percentage of general government expenditure on health; out-of-pocket expenditure – as a percentage of private expenditure on health; and life expectancy. There were strong positive correlations (P < 0.05 for each) between the same time interval and private expenditure on health – as a percentage of total expenditure on health; the estimated annual number of road traffic deaths per 100 000 inhabitants; and the Gini coefficient.
There were strong inverse correlations (P < 0.05 for each) between the interval from admission to surgery and adult literacy, life expectancy, total expenditure on health per capita, government expenditure on health per capita, the human development index and GDP per capita. There were strong positive correlations (P < 0.05 for each) between the same time interval and the numbers of infant deaths per 1000 live births; deaths among children aged less than 5 years per 1000 live births; and the estimated annual number of road traffic deaths per 100 000 inhabitants.
There were strong inverse correlations (P < 0.05 for each) between the surgery–discharge interval, the number of infant deaths per 1000 live births, the number of deaths among children aged less than 5 years per 1000 live births, private expenditure on health – as a percentage of total expenditure on health, the modelled number of road traffic deaths, and the Gini coefficient. There were strong positive correlations (P < 0.05 for each) between the same interval and life expectancy, general government expenditure on health – as a percentage of total expenditure on health – and social security expenditure on health – as a percentage of general government expenditure on health.
This study evaluated whether the intervals between the main stages in the treatment of an isolated femur fracture could be used as indicators of the quality of the clinical process in trauma systems. Indicators of clinical quality need to be well defined, related to optimal patient care, feasibly measured, and able to show improvement.21 The time intervals that we investigated met all of these criteria. There were significant differences – in the timing of the treatment of femur fractures – between the trauma systems in high-income countries and those in low-or middle-income countries. The three types of time interval studied appeared to be valid quality indicators – since each type correlated with several variables related to the allocation of health-care resources. They could be reliably and easily measured – with the minimal use of resources – even in low-income countries.
The interval from injury to admission is a useful tool for examining one of the major defining variables in trauma survival: the time it takes to reach a hospital. In trauma, the general goal is to have the patient assessed within the first hour post-injury – the so-called “golden hour”.34 Pre-hospital time has been examined in various forms for other traumatic injuries in previous studies.5,7,10,26 It has been shown that – for high-velocity trauma – the earlier a patient receives initial treatment, the better the outcome.34–36
The longest times from injury to admission occurred in hospital systems in low-income countries that did not have adequate resources available for the initiation of emergency medical services in the field. In Kenya and the United Republic of Tanzania, for example, patients were more likely to be transported to the hospital via a privately owned car or taxi (50% of Kenyan patients and 94% of the Tanzanian) than by emergency medical services. In a previous study in Kenya, it was reported that trauma patients were transported to the hospital by individuals who had witnessed their injury – so-called “good Samaritans” (76%), family or friends (16%) or the police (6%).37 Private expenditure on health as a percentage of total expenditure on health – which was significantly positively correlated with the interval from injury to admission – is a measure of the private sector contribution to total spending on health care. If private expenditure on health as a percentage of total expenditure on health is high, the government funding available for mobile emergency medical services is likely to be lower and the interval between injury and admission is likely to be relatively long (Table 5).
The Gini coefficient is commonly used to measure inequality of wealth within a population. A Gini coefficient of 0 indicates perfect equality – everybody with the same wealth – whereas a value of 1 indicates perfect inequality – a single individual having all the wealth.38 In our study, the Gini coefficient was found to be positively correlated to the injury–admission interval but negatively correlated with the interval from surgery to discharge. This may be explained by the financial dynamics of the individual trauma and health-care systems. For example, when wealth is concentrated, governments are less likely to allocate resources for pre-hospital care and trauma cases become responsible for their own transport to a hospital. Additionally, when individual expenditure is high within a hospital, relatively impoverished patients are highly motivated to leave a hospital as soon as possible, to minimize the out-of-pocket payments for their hospitalization.
The interval between hospital admission and definitive treatment is typically affected by the personnel and hard goods available as resources, and by the payment system involved. In the present study, this interval was found to be positively correlated with private health expenditure as a percentage of total health expenditure – regardless of the type of health facility. Catastrophic personal expenditures on health care have been shown to cause impoverishment, especially when the patient is one of the main sources of household income. In low- and middle-income countries, the main income earners tend to be young adult males.39–42 In Kenya and the United Republic of Tanzania, we found that patients and their families and friends had had to sell livestock and assets to pay for implants and hospital care. In a previous study in Kenya, 60% of the patients investigated had had to approach relatives for help in paying hospital bills, and 15% had had to pay “up front” before they had received any care.37
The data that we analysed showed a significant inverse correlation between the in-hospital resources that were available and the interval from admission to surgery. In low- and middle-income countries, most patients must pay for resources such as implants before surgery. The acquisition of the necessary implants can be a time-consuming process that delays surgical care. In most hospitals in high-income countries, however, implants and other in-hospital resources are readily available for patient care. In the hospitals in low- and middle-income countries that we investigated – even at the highest level of tertiary care, with 24-hour orthopaedic, emergency and surgical availability – in-hospital resources such as fluoroscopy equipment and anaesthesiologists were relatively scarce. In these hospitals, the absence or scarcity of fluoroscopy machines and other essential equipment meant that femoral nailing was performed as an open procedure much more often than in the study hospitals in high-income countries. Open reduction has become relatively rare in high-income countries because it is associated with higher rates of infection and non-union.43–47 In Kenya and the United Republic of Tanzania, where most orthopaedic surgeons are concentrated in the large urban referral centres, peripheral hospitals seldom treat femur fractures surgically.
The interval from surgery to discharge represents the time spent in in-hospital recovery after surgical intervention. In the present study, there was a high positive correlation between this interval and both government and social security expenditures on health – as proportions of the total expenditure on health. Trauma patients often need sub-acute care before they are discharged and such care may not be readily available in areas with minimal resources. There were strong inverse correlations between the interval from surgery to discharge and private expenditure on health – as a percentage of total expenditure on health; the modelled number of road traffic deaths; and the Gini coefficient. It seems likely that a patient who is facing relatively high out-of-pocket expenditure for hospitalization will be particularly eager to be discharged.
Our study had several limitations. It was difficult to distinguish cases of isolated femur fracture from cases with multiple traumas that included femur fracture, and this may have resulted in selection bias and misclassification. We tried to minimize this problem by interviewing patients – when possible – and searching each patient’s chart for other injuries or diagnoses. We studied far more patients from the United States than from any low- or middle-income country and the number of patients that we studied in each hospital in a low- or middle-income country was relatively small, although we still detected strong correlations. The hospitals that we investigated were all tertiary-care institutions and in most low- and middle-income countries there are many peripheral hospitals for each such tertiary-care institution. The peripheral rural hospitals are particularly likely to have low resources27 and patients attending such hospitals with fractured femurs are more likely to be treated with traction than with intramedullary nailing. Our data probably therefore underestimate the mean length of time it would take a patient with a fractured femur to flow through a trauma system in a low- or middle-income country. Patients with femur fractures in peripheral hospitals may be in traction for many weeks, never receive any surgical intervention and have poorer outcomes – at a higher overall cost to the trauma system – than patients treated by intramedullary nailing.24 In an evaluation of trauma care in a rural setting near Lake Naivasha, Kenya, it was found that half of the patients with fractured femurs were treated with traction or closed treatment.48 Furthermore, we do not know if the hospital patients that we investigated received care elsewhere before presenting to a tertiary-care institution.
In conclusion, the intervals between injury and admission, admission and surgery, and surgery and discharge for patients with fractured femurs were all easily measurable and highly correlated to known, accessible and quantifiable country data on health and economics. The strengths of the observed correlations suggest that the intervals can be used as valid clinical indicators of the quality of trauma systems and as guides to resource allocation efforts. Further work will include a larger multicentre study as well as a study analysing the rate of capture of different traumatic injuries.
We thank our many generous colleagues at the study hospitals for their assistance with the data collection.
This research was funded in part by a grant from the Orthopaedic Research and Education Foundation.
- Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al., et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012; 380: 2095-128 http://dx.doi.org/10.1016/S0140-6736(12)61728-0 pmid: 23245604.
- Murray CJL, Vos T, Lozano R, Naghavi M, Flaxman AD, Michaud C, et al., et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the global burden of disease study 2010. Lancet 2012; 380: 2197-223 http://dx.doi.org/10.1016/S0140-6736(12)61689-4 pmid: 23245608.
- Mock C, Joshipura M, Quansah R, Arreola-Risa C. Advancing injury prevention and trauma care in North America and globally. Surg Clin North Am 2007; 87: 1-19 http://dx.doi.org/10.1016/j.suc.2006.09.017 pmid: 17127120.
- Evans C, Howes D, Pickett W, Dagnone L. Audit filters for improving processes of care and clinical outcomes in trauma systems. Cochrane Database Syst Rev 2009; 4: CD007590 pmid: 19821431.
- London JA, Mock CN, Quansah RE, Abantanga FA, Jurkovich GJ. Priorities for improving hospital-based trauma care in an African city. J Trauma 2001; 51: 747-53 http://dx.doi.org/10.1097/00005373-200110000-00021 pmid: 11586170.
- Mann NC, Mullins RJ, MacKenzie EJ, Jurkovich GJ, Mock CN. Systematic review of published evidence regarding trauma system effectiveness. J Trauma 1999; 47: S25-33 http://dx.doi.org/10.1097/00005373-199909001-00007 pmid: 10496607.
- Mock C, Quansah R, Krishnan R, Arreola-Risa C, Rivara F. Strengthening the prevention and care of injuries worldwide. Lancet 2004; 363: 2172-9 http://dx.doi.org/10.1016/S0140-6736(04)16510-0 pmid: 15220042.
- Pasquale MD. Outcomes for trauma: is there an end (result) in sight? J Trauma 2008; 64: 60-5 http://dx.doi.org/10.1097/TA.0b013e31815eb112 pmid: 18188100.
- Spiegel DA, Gosselin RA, Coughlin RR, Joshipura M, Browner BD, Dormans JP. The burden of musculoskeletal injury in low and middle-income countries: challenges and opportunities. J Bone Joint Surg Am 2008; 90: 915-23 http://dx.doi.org/10.2106/JBJS.G.00637 pmid: 18381331.
- Stelfox HT, Bobranska-Artiuch B, Nathens A, Straus SE. Quality indicators for evaluating trauma care: a scoping review. Arch Surg 2010; 145: 286-95 http://dx.doi.org/10.1001/archsurg.2009.289 pmid: 20231631.
- Stelfox HT, Bobranska-Artiuch B, Nathens A, Straus SE. A systematic review of quality indicators for evaluating pediatric trauma care. Crit Care Med 2010; 38: 1187-96 http://dx.doi.org/10.1097/CCM.0b013e3181d455fe pmid: 20154596.
- Bickler SW, Spiegel DA. Global surgery – defining a research agenda. Lancet 2008; 372: 90-2 http://dx.doi.org/10.1016/S0140-6736(08)60924-1 pmid: 18582930.
- Weiser TG, Regenbogen SE, Thompson KD, Haynes AB, Lipsitz SR, Berry WR, et al., et al. An estimation of the global volume of surgery: a modelling strategy based on available data. Lancet 2008; 372: 139-44 http://dx.doi.org/10.1016/S0140-6736(08)60878-8 pmid: 18582931.
- Hofman K, Primack A, Keusch G, Hrynkow S. Addressing the growing burden of trauma and injury in low- and middle-income countries. Am J Public Health 2005; 95: 13-7 http://dx.doi.org/10.2105/AJPH.2004.039354 pmid: 15623852.
- Hughes BB, Kuhn R, Peterson CM, Rothman DS, Solórzano JR, Mathers CD, et al., et al. Projections of global health outcomes from 2005 to 2060 using the International Futures integrated forecasting model. Bull World Health Organ 2011; 89: 478-86 http://dx.doi.org/10.2471/BLT.10.083766 pmid: 21734761.
- Naddumba EK. Musculoskeletal trauma services in Uganda. Clin Orthop Relat Res 2008; 466: 2317-22 http://dx.doi.org/10.1007/s11999-008-0369-2 pmid: 18629599.
- O’Neill B, Mohan D. Reducing motor vehicle crash deaths and injuries in newly motorising countries. BMJ 2002; 324: 1142-5 http://dx.doi.org/10.1136/bmj.324.7346.1142 pmid: 12003889.
- Peden M. Global collaboration on road traffic injury prevention. Int J Inj Contr Saf Promot 2005; 12: 85-91 http://dx.doi.org/10.1080/15660970500086130 pmid: 16156532.
- Peden M, Toroyan T. Road traffic injuries in South Asia: national and organisational policy responses. J Coll Physicians Surg Pak 2004; 14: 722-5 pmid: 15610629.
- Mock C. WHA resolution on trauma and emergency care services. Injury prevention. Journal of the International Society for Child and Adolescent Injury Prevention. 2007; 13: 285-6 http://dx.doi.org/10.1136/ip.2007.016485.
- Wollersheim H, Hermens R, Hulscher M, Braspenning J, Ouwens M, Schouten J, et al., et al. Clinical indicators: development and applications. Neth J Med 2007; 65: 15-22 pmid: 17293635.
- Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, et al., et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012; 380: 2163-96 http://dx.doi.org/10.1016/S0140-6736(12)61729-2 pmid: 23245607.
- Wolinsky PR, McCarty E, Shyr Y, Johnson K. Reamed intramedullary nailing of the femur: 551 cases. J Trauma 1999; 46: 392-9 http://dx.doi.org/10.1097/00005373-199903000-00007 pmid: 10088839.
- Gosselin RA, Heitto M, Zirkle L. Cost-effectiveness of replacing skeletal traction by interlocked intramedullary nailing for femoral shaft fractures in a provincial trauma hospital in Cambodia. Int Orthop 2009; 33: 1445-8 http://dx.doi.org/10.1007/s00264-009-0798-x pmid: 19437019.
- Starr AJ, Hay MT, Reinert CM, Borer DS, Christensen KC. Cephalomedullary nails in the treatment of high-energy proximal femur fractures in young patients: a prospective, randomized comparison of trochanteric versus piriformis fossa entry portal. J Orthop Trauma 2006; 20: 240-6 http://dx.doi.org/10.1097/00005131-200604000-00002 pmid: 16721238.
- Henry JA, Reingold AL. Prehospital trauma systems reduce mortality in developing countries: a systematic review and meta-analysis. J Trauma Acute Care Surg 2012; 73: 261-8 http://dx.doi.org/10.1097/TA.0b013e31824bde1e pmid: 22743393.
- Penoyar T, Cohen H, Kibatala P, Magoda A, Saguti G, Noel L, et al., et al. Emergency and surgery services of primary hospitals in the United Republic of Tanzania. BMJ Open 2012; 2: e000369 http://dx.doi.org/10.1136/bmjopen-2011-000369 pmid: 22307096.
- Tinney MJ, Chiodo A, Haig A, Wiredu E. Medical rehabilitation in Ghana. Disabil Rehabil 2007; 29: 921-7 http://dx.doi.org/10.1080/09638280701240482 pmid: 17577726.
- Guidelines for essential trauma care. Geneva: World Health Organization; 2012.
- American College of Surgeons [Internet]. National Trauma Data Bank® (NTDB). Chicago: ACS; 2013. Available from: http://www.facs.org/trauma/ntdb/ [accessed 27 September 2013].
- Linn S. Sorting variables for each case: a new algorithm to calculate injury severity score (ISS) using SPSS-PC. Stat Med 1996; 15: 933-7 http://dx.doi.org/10.1002/(SICI)1097-0258(19960415)15:7/9<933::AID-SIM262>3.0.CO;2-5 pmid: 8861162.
- World Bank list of economies – November 2011. Washington: World Bank; 2011.
- World Health Organization data and statistics. Geneva: World Health Organization; 2011.
- Advanced trauma life support student course manual. Chicago: American College of Surgeons; 2012.
- Tien HCN, Jung V, Pinto R, Mainprize T, Scales DC, Rizoli SB. Reducing time-to-treatment decreases mortality of trauma patients with acute subdural hematoma. Ann Surg 2011; 253: 1178-83 http://dx.doi.org/10.1097/SLA.0b013e318217e339 pmid: 21494125.
- Fatovich DM, Phillips M, Jacobs IG, Langford SA. Major trauma patients transferred from rural and remote Western Australia by the Royal Flying Doctor Service. J Trauma 2011; 71: 1816-20 http://dx.doi.org/10.1097/TA.0b013e318238bd4c pmid: 22027890.
- Macharia WM, Njeru EK, Muli-Musiime F, Nantulya V. Severe road traffic injuries in Kenya, quality of care and access. Afr Health Sci 2009; 9: 118-24 pmid: 19652745.
- González Abril L, Vaelasco Morente F, Gavilán Ruiz J, Sánchez-Reyes Fernández L. The similarity between the square of the coeficient of variation and the Gini index of a general random variable. J Quant Methods Econ Bus Adm 2010; 10: 5-18.
- Gopalan SS, Das A. Household economic impact of an emerging disease in terms of catastrophic out-of-pocket health care expenditure and loss of productivity: investigation of an outbreak of chikungunya in Orissa, India. J Vector Borne Dis 2009; 46: 57-64 pmid: 19326709.
- Kawabata K, Xu K, Carrin G. Preventing impoverishment through protection against catastrophic health expenditure. Bull World Health Organ 2002; 80: 612 pmid: 12219150.
- Van Damme W, Meessen B, Por I, Kober K. Catastrophic health expenditure. Lancet 2003; 362: 996-author reply 997 http://dx.doi.org/10.1016/S0140-6736(03)14375-9 pmid: 14511936.
- Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, Murray CJ. Household catastrophic health expenditure: a multicountry analysis. Lancet 2003; 362: 111-7 http://dx.doi.org/10.1016/S0140-6736(03)13861-5 pmid: 12867110.
- Ebong WW. Open medullary nailing of the femur: Ibadan experience. East Afr Med J 1984; 61: 393-8 pmid: 6510317.
- Eid AM. Aetiological factors in infection following open retrograde intramedullary nailing of the femur. Postgrad Med J 1979; 55: 797-9 http://dx.doi.org/10.1136/pgmj.55.649.797 pmid: 392488.
- Harper MC. Fractures of the femur treated by open and closed intramedullary nailing using the fluted rod. J Bone Joint Surg Am 1985; 67: 699-708 pmid: 3997922.
- Savastano AA, Cadena E. A comparative study of the treatment of fractures of the shaft of the femur by traction and open fixation: traction indicated for patients under sixteen, with intramedullary nailing reserved for patients over twenty-two. R I Med J 1971; 54: 144-6 pmid: 5279721.
- Schatzker J. Open intramedullary nailing of the femur. Orthop Clin North Am 1980; 11: 623-31 pmid: 7413178.
- Otieno T, Woodfield JC, Bird P, Hill AG. Trauma in rural Kenya. Injury 2004; 35: 1228-33 http://dx.doi.org/10.1016/j.injury.2004.03.013 pmid: 15561111.