Missed injuries: A Ugandan experience

Article Outline

• Summary
• Introduction
• Method
• Results
• Discussion
• Conclusion
• References
• Copyright

Summary 

Background

Missed injuries (MIs) have been noted worldwide in all trauma centres that have studied them, and they are a significant cause of patient morbidity and mortality.

Objective

To establish the prevalence, contributing factors and short-term outcome of missed injuries in cases of multiple and major trauma.

Method

Longitudinal prospective study involving 403 patients over 5 months.

Results

Missed injuries were discovered in 78 cases (prevalence 19.4%). Contributing factors included incomplete assessment (52.5%), radiological errors, surgical failures and patient's arrival time. The most affected body regions were the head and neck, extremities and pelvic girdle and contents; in the abdomen, 49.1% of injuries were missed. Among the 28 deaths in the study, 21 occurred in cases with missed injuries, and 13 (62%) of these deaths were directly attributable to missed injuries (R²=12.5, p=0.0001, 95% CI 5.5–28.35).

Conclusion

There is need for improvement in patient assessment and monitoring, in efficiency of the trauma team, and for staff redistribution to address the increase in night arrivals.

Keywords: Missed injuries, Major trauma, Multiple trauma, Short-term outcome

Back to Article Outline

Introduction 

Road traffic crashes (RTCs) and interpersonal violence continue to contribute to an increased incidence of injuries in Uganda.¹⁰, ¹¹ The care of these injured patients is a consistent challenge to any accident and emergency (A&E) unit. The challenge is even more conspicuous when dealing with the multiply injured, for whom resuscitation, diagnosis and therapy have to proceed simultaneously. These numerous and complex tasks come with the risk of missing some injuries, which contributes significantly to morbidity and mortality. In this study, injury was defined as a missed injury (MI) if it was not detected or documented before onward transfer from the A&E holding ward. The study set out to determine the prevalence, causes and short-term outcome of MIs in Mulago Hospital, Kampala. The findings would be used to address causes of MIs as well as to improve patient care.

Back to Article Outline

Method 

This longitudinal prospective study was conducted over 5 months (August 2004 to January 2005), in the A&E unit and surgical firms of Mulago Hospital, which is a teaching and national referral hospital with 1500 beds.

On arrival in the A&E unit, an injured patient is normally received by the triage nurse who, whenever possible, records baseline information before the medical officer on duty carries out the primary survey. Resuscitation is initiated at this point (including operative resuscitation if needed). On completion of the primary survey, the patient is transferred to the A&E holding ward, where the secondary survey is performed by the medical officer and senior house officer of the admitting firm. A tertiary survey is later carried out by the firm registrar or consultant during the next round, before transfer to the admitting firm.

All patients seen in A&E with multiple and or major injuries were eligible for recruitment to our study. We excluded those who died without initial assessment, unaccompanied minors and unconscious adults without next of kin to assent. We consecutively recruited 403 patients (minimum sample size=284), who were examined daily either until discharge or for up to 30 days. Their charts were reviewed and any unaddressed complaints were investigated.

Variables recorded included: demographic characteristics; times of injury; arrival and receipt of care; mechanism and site of injury; injury scores, including the Abbreviated Injury Score (AIS), New Injury Severity Score (NISS) and Glasgow Coma Score (GCS);¹², ¹⁶ investigations and results; designation of attending officers; and morbidity and mortality. In the event of death, a post-mortem examination was carried out. Because the investigators were themselves clinicians, if a missed injury was revealed they intervened or caused intervention. Therefore, certain outcomes that would have otherwise been observed in a retrospective study design were not observed. Permission to carry out this study was duly obtained from the relevant authorities. Data were analysed using EPIINFO 2000, SPSS 10.0 for Windows and Stata 8.0; a p-value of <0.05 was taken as statistically significant.

Back to Article Outline

Results 

We found that 78 patients had MIs (prevalence 19.4%), with an average of 1.7 MIs per patient. The patients’ characteristics are documented in Table 1.

Table 1. Comparison^a of patients’ characteristics in the two groups

Parameter	No MI (n=325)	MI (n=78)	p-value
Mean age in years	29	29	NS
Male gender (%)	270 (83.1%)	61 (78.2%)	NS
Female gender (%)	55 (16.9%)	17 (21.8%)	NS
GCS mean (median)	14 (15)	12 (13)	0.001
NISS mean (median)	15 (14)	24 (25)	<0.001
Injuries per patient (mean)	2.4	3.9
Daytime arrival (%)	197 (89.4%)	26 (11.6%)	<0.001
Night arrival (%)	128 (71%)	52 (29%)	<0.001
Outcome
Discharged (%)	254 (78.1%)	36 (46.2%)
Still in hospital (%)	64 (19.7%)	21 (26.9%)
Mean days in hospital	13	16
Died (%)	7 (2.2%)	21 (26.9%)

MI, missed injuries; NS, not significant; GCS, Glasgow Coma Score; NISS, New Injury Severity Score.

Of the 403 patients recruited, 388 (96.3%) sustained 2 or more injuries; of these 388, 76 had missed injuries, of whom 71 (91%) sustained major injuries (NISS>15).

Using the GCS, 48 (61.5%) had minor head injuries (GCS 13–15), 14 (18%) had moderate head injuries (GCS 9–12) and 16 (20.5%) had severe head injuries (GCS<9).

The mechanism of injury was RTC in 57 (73.1%) cases, fall in 7 (9%), missile injuries in 7 (9%), assault in 5 (6.4%) and other in 2 (2.5%); 28 (49%) of the RTCs were associated with public transport by motor cycle, as cyclist, passenger or pedestrian.

We found that 52.5% of the injuries occurred outside the body area of clinical focus (type I), 41.1% occurred within the area of clinical focus or surgical attention (type II) and 6.4% were a result of the surgeon's decision to abbreviate surgery (type III).⁷ The MIs were also grouped according to AIS² body regions, as shown in Table 2.

Table 2. Location, nature and number of missed injuries

AIS, Abbreviated Injury Score; MIs, missed injuries; N, number of injuries in study population (403) in the same AIS body region.

Multiple factors contributed to the occurrence of MIs, the most frequent being incomplete assessment of the AIS body areas (accounting for 52.5% of MIs). Among type I MIs,⁷ this occurred 33 times and was followed by failure to assign significance to an apparently superficial injury overlying, for example, a ruptured spleen or urethral disruption. The occurrence of an MI was also significantly influenced by the patient's arrival time (Fig. 1) and the seniority of the attending officer (Table 3). We noted that 52 patients with missed injuries had been admitted during the night, compared with 26 admitted during the day. The mean waiting time between arrival and assessment was 66min during the night period and 28min during the day period.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Patients’ arrival time. Red indicates missed injuries, green indicates no missed injuries. Arrival time: (A) 08:00–11:59h; (B) 12:00–15:59h; (C) 16:00–19:59h; (D) 20:00–23:59h; (E) 00:00–03:59h; (F) 04:00–07:59h.

Table 3. Most senior medical officer attending at time of diagnosis for definitive care

Designation	No MIs (n=325)	MIs (n=78)	p-value	R2: MIs/no MIs
Consultant	200 (61.5%)	18 (23.1%)	<0.001	0.25
MOSG	100 (30.8%)	19 (24.4%)	0.265	0.776
SHO	22 (6.8%)	33 (42.3%)	<0.001	4.64
MO/intern.	3 (0.9%)	8 (10.2%)	<0.001	4.07
Total	325 (100%)	78 (100%)

MIs, missed injuries; MOSG, medical officer, special grade; SHO, senior house officer; MO, medical officer.

Of the patients who did not have MIs, 61.5% had been reviewed by a consultant surgeon. This is in sharp contrast to the group with MIs, among whom 23.1% had been reviewed by a consultant surgeon. Other factors included altered consciousness (6 cases), radiological error (17), surgical failure (surgery ill-timed or inadequate) (7) and inadequate patient review (3).

A multivariate regression analysis showed that the severity of the injuries (NISS), patient's time of arrival and seniority of the attending officer had significant modifying effects on each other and on the occurrence of a missed injury (p=0.001, R²=0.47). GCS did not affect this relationship significantly. MIs were detected by the consultant surgeon and firm in 40 of the 78 cases, in 31 by the investigating team (the principal investigator and assistant) and in 7 by the radiologist. These MIs were found by detailed examination in 38 cases (48.6%) and were confirmed by laparotomy in 2 cases; they were diagnosed on grounds of no progress in 16 cases, on review of misinterpreted radiographs in 8 and at autopsy in 14. Mean delay in diagnosis of MI was 3 days (range 1–14 days).

The outcomes of the injuries are shown in Table 1. Mean duration of care before death was 3 days (range 1–9 days). Post-mortem examination was carried out in 26 cases (the remaining 2 examinations could not be performed because of religious beliefs). The majority of fatal injuries were located in the head and neck and in the thorax and abdomen.

Back to Article Outline

Discussion 

The fundamental principles of the ATLS (Advanced Trauma Life Support)¹ are widely known, taught and practised in many trauma centres, including Mulago Hospital. Following the ATLS principles, all patients pass through primary and secondary surveys and a definitive care phase. However, it is recognised that a number of injuries still escape detection. In Mulago Hospital, patients pass through the above three phases within 24h of arrival, before onward transfer from the holding ward to a general surgical ward.

This study examines a problem not previously investigated at our centre or any other hospital in the country. The design was prospective and longitudinal, providing a more accurate picture (compared with a retrospective design) of the extent of the problem. Equally important is the fact that it was possible to note different errors during the patient management process and to cause intervention in some cases. We paid particular attention to clinically significant injuries; minor injuries were taken into account when they accompanied the former. The findings of this study therefore identify clinically significant injuries missed by tertiary survey in other studies.⁶, ⁹, ¹³, ¹⁴

The missed injury rate of 19.4% in Kampala is high compared with that reported from Toronto (8.1%),³ Odense (8.1%),⁸ Sydney (16%)¹³ and Rhode Island (1.5–3.4%),¹⁴ and these patients had a higher mean NISS and number of injuries (Table 1).

Incomplete assessment of the AIS² body region accounted for over 52.5% of our MIs. As noted, our MI distribution by AIS body region was similar to that of other centres but, in particular, 49.1% of injuries to abdominal and pelvic contents were missed (Table 2). We found that this could be attributable to inadequacy of examination and investigation of the injured abdomen. Wangoda¹⁵ found the accuracy of diagnostic peritoneal lavage in our centre to be 97.6%, with a sensitivity of 100% and specificity of 96.4%. However, none of the patients studied underwent diagnostic peritoneal lavage, and it was difficult to obtain an abdominal ultrasound (US) examination outside regular working hours at our hospital.

Soon after this study, an ultrasound machine became available to our A&E unit. Arrangements are being made to train the appropriate A&E staff in focused abdominal sonography in trauma (FAST) techniques.

A repeated detailed examination is of paramount importance in the prevention of MIs and alone detected these in 38 (48.7%) of our study patients. In our resource-limited setting, clinicians need to develop, retain and rely on clinical acumen. There is need for mandatory documented secondary and tertiary surveys, using appropriate tools. The A&E unit currently runs regular interdisciplinary trauma team training and retraining of staff, to standardise care offered according to ATLS principles. This should quickly build a dedicated trauma team whose zeal will address the gaps present in patient assessment and care, particularly of the severely injured.

Overall, unpublished data from our institution indicate that the majority of trauma victims arrive during the night. In this study, the number of people in the study population with multiple or major injuries arriving during the day (08:00h–19:59h) or night period (20:00h–07:59h) were comparable (55.4% versus 44.6%), but injuries were missed in 11.6% of those attending during the day compared with 29% of night arrivals (Table 1). This was also found true by Janjua at Liverpool Hospital, New South Wales.⁹ A breakdown of the arrival times into 4-h intervals established that 51.3% of the patients (40 of 78) with MIs had reached A&E between 16:00h and 00:00h. This coincides with the daily duty changeovers at 16:00h and 21:00h. Mean waiting time from arrival to assessment during the day period was 28min compared with 66min in the night period.

The evening and night hours are also challenging because then auxiliary units (radiology, laboratories) scale down their operations. Furthermore, senior surgical staff, whom we found to be pivotal in diagnosis of MIs, were unlikely to be present at night unless called upon for difficult cases. In this study, only 18 (23.1%) of the patients who had MIs had been reviewed by a consultant surgeon by the time the MI was diagnosed and definitive treatment instituted (Table 3).

In a setting such as ours, where staff shortage is an incessant challenge, distribution of the few staff available needs to be designed to reduce such gaps, and there is now an ongoing debate within the department to address this issue.

Radiological errors occurred in 17 (22%) cases and included the choice of investigation, the views taken, limitations of the technique chosen and interpretation of radiographs. Radiographs that are returned soon after the secondary survey are interpreted by the medical officer, senior house officer or surgical registrar present. Although these clinicians have acquired skills in interpretation of radiographs, it is inevitable that some abnormalities pass unnoticed. An experienced radiologist's report can take up to a few days to obtain. Currently, there is a surgical–radiological conference once a week, which addresses most issues in retrospect. A daily formal conference of the two teams would prove very useful in revealing otherwise unrecognised issues.

Outcomes of care are shown in Table 1. The patients with MIs stayed longer in hospital (mean stay 16 days versus 13 days for those without MIs). This was because the injuries were not only diagnosed late, but also because more time was required to adequately treat them and their resultant complications, such as shock, peritonitis and vascular injury. The prolonged stay could also have been due to the severity of the injuries in this group as evidenced by the higher mean NISS (Table 1).

In a few cases, early recognition of the MI would have prevented unfavourable outcomes such as loss of a finger due to ischaemic gangrene, pain and meningitis in a case with a missed open head injury. In some cases, however, early recognition of the MI would not have necessarily altered the outcome, e.g. facial palsy, foot drop, expressive aphasia, fractured femur. Nonetheless, an injury unnoticed, however trivial, is likely to cause the patient varying degrees of distress and affects the doctor–patient relationship. The most noteworthy outcome was death as a result of missed injuries; 21 (75%) of the 28 deaths in this study occurred in cases of MI. In 13 (46.4%) of the fatalities, the injuries found at post-mortem examination could have directly caused the death. Timely surgical intervention, therefore, might have significantly altered the course of events. In one such case, a person with a grade IV kidney injury, perforated sigmoid colon and fractured ribs, also had an ominously eye-catching deep laceration to the left thigh. The attending clinician, in the operating room, carried a surgical toilet and sutured the wound on the thigh. The patient deteriorated and died 4h later while in the holding ward. For the remaining 8 patients with missed injuries who died, a timely diagnosis would not have necessarily altered the outcome observed, because of severity of the injuries.

The mortality figures in this study are similar to those reported from Houston, Texas,⁵ in which 64% of deaths were directly attributable to the missed injuries, but compare quite inauspiciously with those from other more recent studies. Buduhan and McRitchie³ in Toronto found 1 death directly attributable to a missed injury among 46 patients. Sung and Kim⁴ in South Korea reported 2 deaths among 12 patients they followed up. Conversely, Janjua and Sugrue⁹ found 25 MIs in 9 patients at autopsy, but none of these MIs contributed to the death.

It is worth noting that post-mortem examinations are very rewarding in the diagnosis of missed injuries, as shown in our study and in other studies cited above. Therefore, the care of a trauma patient should not be considered complete until a review of care offered has been carried out or, in the event of death, a post-mortem examination has been offered. Currently our A&E unit has instituted a weekly audit, and this should go a long way towards addressing the circumstances surrounding such events.

Back to Article Outline

Conclusion 

The high prevalence of clinically significant missed injuries in this institution calls for deliberate action. Strategies should aim at training and retraining trauma team members in adequate patient evaluation based on ATLS principles, making a consultant available on 24-h basis, routine surgical audit (including post-mortem examination) and reorganisation of the trauma team and working day to increase the presence of night staff. This will ensure that the next injury patient is not another morbidity or mortality statistic.

We recommend that, in the near future, a similar study should be performed to investigate the impact of the various interventions being instituted to improve the quality of care of the injured patient.

Back to Article Outline

References 

1. American College of Surgeons Committee on Trauma. Initial assessment and management. ATLS reference manual. Chicago: American College of Surgeons; 1994. p. 17–37.
   • View In Article
2. Association for the Advancement of Automotive Medicine. The Abbreviated Injury Scale, revision; 1990.
   • View In Article
3. Buduhan G, McRitchie DI. Missed injuries in patients with multiple trauma. J Trauma. 2000;49:600–605
   • View In Article
   • MEDLINE
4. Chun KS, Kong HK. Missed injuries in abdominal trauma. J Trauma. 1996;41:276–278
   • View In Article
   • MEDLINE
5. Enderson BL, Maull KI. Missed injuries. The trauma surgeon's nemesis. Surg Clin North Am. 1991;71:399–418
   • View In Article
   • MEDLINE
6. Enderson BL, Reath DB, Meadors J, et al. The tertiary trauma survey: a prospective study of missed injury. J Trauma. 1990;30:666–670
   • View In Article
   • MEDLINE
7. Hirschberg A, Wall JM, Allen MK, et al. Causes and patterns of missed injuries in trauma. Am J Surg. 1994;168:299–303
   • View In Article
   • MEDLINE
   • CrossRef
8. Houshian S, Larsen MS, Holm C. Missed injuries in a level 1 trauma centre. J Trauma. 2002;52:715–719
   • View In Article
   • MEDLINE
9. Janjua KJ, Sugrue M, Deane SA. Prospective evaluation of early missed injuries and the role of tertiary survey. J Trauma. 1998;44:1000–1007
   • View In Article
   • MEDLINE
10. Kirya FM, Kijjambu S, Ezati IA. Outcome of major trauma at Mulago Hospital in Uganda. Assessment using the TRISS methodology. East Central Afr J Surg. 2002;7:59–62
    • View In Article
11. Kobusingye OC, Guwatude D, Owor G, et al. Citywide trauma experience in Kampala, Uganda: a call for intervention. Inj Prev. 2002;8:133–136
    • View In Article
    • MEDLINE
    • CrossRef
12. Lefering R. Trauma score systems for quality assessment. Eur J Trauma. 2002;28:52–63
    • View In Article
13. Soundappan SVS, Andrew JAH, Daniel TC. Role of an extended tertiary survey in detecting missed injuries in children. J Trauma. 2004;57:114–118
    • View In Article
    • MEDLINE
14. Walter LB, Harrington DT, Cioffi WG. Implementation of a tertiary trauma survey decreases missed injuries. J Trauma. 2003;54:38–44
    • View In Article
    • MEDLINE
15. Wangoda R, Upoki A, Owori FN, Kawooya MG. The value of diagnostic peritoneal lavage in the initial assessment of blunt abdominal trauma in Mulago Hospital. MMed dissertation, Makerere University Medical School, Kampala; 2001.
    • View In Article
16. Balogh ZJ, Varga E, Tomka J, et al. The new injury severity score is a better predictor of extended hospitalization and intensive care unit admission than the injury severity score in patients with multiple orthopaedic injuries. J Orthop Trauma. 2003;17:508;Makerere University Medical School 12
    • View In Article
    • MEDLINE
    • CrossRef