- Email: [email protected]
Severe and multiple trauma in older patients; incidence and mortality
Injury, Int. J. Care Injured 40 (2009) 362–367
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Severe and multiple trauma in older patients; incidence and mortality P.V. Giannoudis a,*, P.J. Harwood a, C. Court-Brown b, H.C. Pape c a
Academic Department Orthopaedic Trauma Surgery, Leeds University, Great George Street, Leeds, West Yorkshire LS1 3EX, UK Edinburgh Orthopaedic Trauma Unit, Royal Infirmary of Edinburgh, Edinburgh, UK c Department of Orthopaedics, University of Pittsburgh Medical Centre, Pittsburgh, USA b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 13 October 2008
Objective: To examine the differences between severely injured older patients (aged over 65 years) compared with similarly injured younger adults in terms of incidence, inpatient mortality and factors predicting outcome. Methods: Data prospectively entered into the Trauma Audit and Research Network (TARN) database from our level I trauma unit over a 5-year period were retrospectively examined, with 3172 patients included in the final analysis. Results: Older patients accounted for 13.8% of those with severe injuries (Injury Severity Score 16 or more) and almost 2% of our trauma admissions overall. High energy injuries were responsible for the majority of these injuries though relatively minor trauma became increasingly important in older patients. Mortality rates in the older patients were more than twice those seen in the adult population (19% in the under 40’s to almost 50% in the over 75’s). Age, Injury Severity Score and Glasgow Coma Score continued to be predictive of mortality in older patients but other factors relevant in younger adults were not. Conclusions: Patients in the older group without physiological derangement on admission were still at a relatively high risk of inpatient mortality. This was in contrast to the younger patients, suggesting that it might be more difficult to predict which older patients might benefit from more aggressive monitoring or treatment. Despite increased mortality in older patients, significant survival rates were achieved even in the oldest. Active treatment should not be withdrawn on the basis of age alone. ß 2008 Elsevier Ltd. All rights reserved.
Keywords: Polytrauma Multiple injury Outcome Mortality Adult Elderly Age
Introduction Background/importance Recent reports of polytrauma in older adults are limited and the subject has previously been given only cursory attention. With an aging, increasingly active older population it is likely that such patients will be seen with increasing frequency.26 Traditional trauma protocols have been established and proven in the treatment of younger accident victims. Older people with diminished physiological reserve, often in association with significant co-morbidities, require special consideration. The distribution of injuries and type of injury mechanism is likely to be different in a population with a high incidence of osteoporosis.21 Older patients can become multiply injured following low energy trauma and these injuries may have worse outcomes.24 Whilst
low energy falls have been reported to account for only 9–11% of injury related deaths in the general population, they comprise more than 50% of traumatic deaths in persons over 65 years old.3,13 Patients with limited mental or physical capacity are also more likely to be involved in accidents as they are slower to identify and respond to dangerous situations.15,16,25 One must also consider the likelihood of a medical emergency such as a myocardial infarction or stroke precipitating an accident, making it necessary to treat this pathology alongside the patient’s injuries. Mental and physical incapacity can make assessment of these patients troublesome. Older patients may become confused and uncooperative during comparatively minor physiological derangements, often compromising their investigation or treatment.5,7 For these reasons it is important to accurately document the pattern of injuries and clinical course of older patients with severe injuries. Goals of the study
* Corresponding author. Tel.: +44 113 2432799. E-mail addresses: [email protected] (P.V. Giannoudis), [email protected] (P.J. Harwood). 0020–1383/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2008.10.016
The aim of this study was to identify and characterise older patients with multiple injuries in our trauma population and
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
compare them to younger patients in terms of their injury patterns, mortality, complications and factors predictive of poor outcome. Materials and methods Data source/study design Data regarding trauma patient care and outcome in our institution is prospectively entered into the Trauma Audit and Research Network (TARN) database. This is an ongoing national study of trauma epidemiology in the United Kingdom.19 The database includes all injured patients arriving alive at the hospital who are then admitted for more than 72 h, or who die in hospital during this initial 72 h period. Patients aged greater than 65 years with isolated fractures of the proximal femur or pubic ramus and patients with simple isolated injuries are excluded. The database includes details of patient demographics, on scene and admission physiology (including pulse, systolic blood pressure, respiratory rate and Glasgow Coma Score), initial resuscitation, inter-hospital transfers and operative intervention. Abbreviated Injury Scale (AIS) injury codes, and the Injury Severity Scores (ISS),4 duration of intensive care unit stay, length of hospital stay or time of death are also documented. This database formed the basis of the study, which therefore represents retrospective analysis of prospectively assembled data. Setting The study included patients admitted to a level one trauma unit, in an urban setting in the United Kingdom. The hospital receives over 7500 admissions per year, directly serves a population of 3.5 million and has access to all medical and surgical specialities on site and acts as a regional tertiary referral centre.
363
Statistical analysis SPSS software was used for statistical analysis (Chicago, IL).18 Parametric and non-parametric tests were used as appropriate to assess continuous variables for significant differences between groups. The Student’s t-test and Mann–Whitney U tests were employed. Dichotomous variables were assessed using a chisquare test. Significant variables were entered into stepwise, multiple logistic regression analysis to determine factors related to outcome. For all outcomes statistical significance was assumed at the p < 0.05 level. The primary outcome used in the study was inpatient mortality from any cause at initial admission. Secondary outcomes included length of inpatient stay, intensive care unit admission, timing of mortality and systemic complications. Results Characteristics of the study group 24,560 patients were entered into the TARN database over the 5-year study period, 13% (3172) were severely injured (ISS equal to or greater than 16). Within the severely injured group, 14% (438) were 65 or older and 16% (505) were under 16. Older patients with severe injuries accounted for 1.8% of the unit’s trauma admissions eligible for entry into the TARN database overall. Fig. 1 shows the frequency distribution of included patients by age. The median age in the younger adult group was 29 (range 16–64 years) compared with 75 (range 65–100 years) in the older group. A significantly greater proportion of patients in the older group were female (42% vs. 22%, p < 0.001). The proportion of female patients remained fairly constant at approximately 22% up to the age 70. Thereafter the proportion of female patients increased dramatically to 41% in those aged 70–80 and 59% in the over 80’s. These increases were statistically significant (p < 0.001).
Selection of patients Analysis of injury severity, hospital stay and treatment Patients admitted to our unit between 1996 and 2001, aged 16 or more, with severe injuries as defined as ISS greater than or equal to 16,4 were identified from the database for inclusion in the study. From this otherwise unselected series, patients were divided into a younger (under 65) and an older (65 or older) group based upon their age at presentation. Further analysis was undertaken dividing the older patient group into 10-year intervals and younger patients into those aged from 16 to 40 and 40 to 65. Data collection and processing Data regarding the above patients were made available from the central database for analysis. The following variables were collected: Demographics and mechanism of injury. Admission physiological variables and details of initial resuscitation. Operative interventions (patients taken to the operating theatre for any treatment). Injury distribution and overall severity (AIS injury codes, and ISS). Duration of intensive care unit admission, hospital stay and time of death. Systemic complications suffered during inpatient treatment including adult respiratory distress syndrome (ARDS), systemic sepsis and multiple organ failure (MOF).
There was no difference in the median ISS between the older and younger adult groups, which was 25 for each. Median length of hospital stay was equivalent at 8 days. If only survivors were considered, the median length of stay was 11 in the younger group and 13 in the older group; this difference was not statistically significant. More patients in the younger group were admitted to ICU and the median length of ICU stay was 4 days (range 0–66) in both groups. 62% of the adult patients underwent operative management during their stay compared with 41% of the older patients (p < 0.001). The proportion of patients receiving operative treatment decreased consecutively with increasing age, from 65% in the under 40’s to 47% in the 65–75’s, 38% in the 75–85’s and 27%
Fig. 1. Age distribution of adult patients admitted between 1996 and 2001 with severe injuries (n = 3172).
364
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
in the over 85’s. In the over 65’s of those operated upon in the first 24 h, 47% underwent neurosurgical, 34% extremity or spinal, 15% abdominal and 7% cardiothoracic interventions. This was not significantly different from the pattern of interventions undertaken in the adult group. No patients over the age of 90 were operated upon. Analysis of injury mechanism and distribution Road traffic accidents (RTA) were the most common cause of injury in each group, though more common in the younger patients (53% vs. 42%, p < 0.001). Falls from height (>2 m) were equally common in each group, whilst minor falls (<2 m) were significantly more common in the older patients (8% vs. 31%, p < 0.001). Younger patients were more likely to be the victims of blunt assault (p < 0.001) or sporting injuries (p = 0.003), both of which were uncommon in either group. Overall anatomic injury distribution was similar between the two groups, 63% of patients had serious (AIS 3 or worse) head, 18% serious thoracic or lower limb injuries and 9% serious abdominal injuries. Younger patients suffered more facial, neck and abdominal injuries overall and more severe (AIS 3 or worse) facial, chest and abdominal injuries when compared with the older patients. Older patients suffered more severe (AIS 3 or worse) external injuries, all of which were severe burns. Analysis of mortality and complications The overall mortality rate was significantly higher in the older patient group (42% vs. 20%, p < 0.001). Fig. 2 shows inpatient mortality by age group. This increased significantly with age from 19% in the under 40’s to almost 50% in the over 75’s. Fig. 3 shows the percentage of those who died in each time period by age group. In both groups, the majority died within the first 24 h, falling progressively with time. However, the overall trend was for a greater proportion of the older patients to die later when compared to the adults. This difference was only statistically significant for the 24–72 h and 2–3-week periods (p < 0.05 and p < 0.001), statistical significance was approached for the two intervening points (p = 0.1). ARDS and MOF were more common in the older than in younger patients (28% vs. 24%, p < 0.05 and 30% vs. 22%, p < 0.001, respectively) whereas systemic sepsis was recorded more commonly in the younger patients (17% vs. 10%, p < 0.01). Considering only the older group, those who died were significantly more likely to have suffered these complications except for sepsis which was not different between the groups (ARDS 34% vs. 23%, p < 0.0001, MOF 35% vs. 14%). Older patients suffering any of these complications were significantly more likely to die compared
Fig. 3. Percentage of deaths occurring in each time period.
with adult patients suffering the same complications (ARDS 51% vs. 23%, sepsis 44% vs. 20%, MOF 59% vs. 32%, all p < 0.0001). Analysis of predictors of outcome on admission Tables 1 and 2 compare various admission variables between survivors and those who died in each group. In the older group, survivors were significantly younger, had a significantly lower ISS, a lower pulse and higher GCS and BP on admission. In the adult group, as well as these parameters, survivors received significantly less fluid in the accident and emergency department and had significantly higher respiratory rates compared with those who died. Cardiac arrest, GCS < 9, BP < 90 on admission and ISS > 25 were significantly associated with increased mortality in both groups. A pulse greater than 90 was associated with increased mortality in the older patients group but not the younger adults. Any head, chest or abdominal injury as well as severe (AIS severity 3 or greater) head, chest, abdominal and spinal injuries were associated with a significantly increased rate of mortality in both groups. Any spinal or external injury was associated with decreased risk of mortality in the adult but not the older group, whilst the risk associated with severe lower limb injuries approached statistical significance in the older but not the younger patients. Other factors considered were not associated with any difference in mortality rates. Table 3 summarises the results of logistic regression analysis carried out on data from the two groups. Variables under consideration were ISS, age, sex, head, chest, abdominal or spinal injuries, cardiac arrest, and admission blood pressure, pulse and respiratory rate, with these physiological values considered both as raw continuous and dichotomous variables (pulse > 90, systolic BP < 90, respiratory rate > 20). For the adult patients, age, female sex, ISS, Glasgow Coma Score, cardiac arrest on admission and systolic blood pressure < 90 were predict as of mortality. The presence of any spinal injury predicted survival (adjusted odds ratio 0.577). For the older patients, only age, ISS and Glasgow Coma Score were significant predictors of mortality. If only patients over the age of 75 are considered, only the ISS and Glasgow Coma Score were significant variables included in the model (adjusted odds ratios 1.07 and 0.83 change per point increase in each score, respectively, both independent predictors of mortality p < 0.001). Discussion Incidence and demographics
Fig. 2. Percentage mortality by age group. *Significant difference between consecutive groups, p < 0.05.
As the population ages and older people become increasingly healthy and active, older victims of multiple trauma will become more frequent. With older people making up 14% of the patients
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
365
Table 1 Characteristics of survivors and those who died during initial hospital admission. Admission variable (median)
Age (years) ISS (16–75) GCS (3–15) Pulse (bpm) Systolic BP (mmHg) Respiratory rate (per minute)
Older group (65 and over, 438 patients)
Younger group (Under 65, 2229 patients)
Survived (254 patients)
Died (184 patients)
Significance
Survived (1783 patients)
Died (446 patients)
Significance
74 20 14 80 140 18
76* 26* 9* 88 130* 18
0.013 0.001 0.001 0.021 0.024 Not significant
33 25 14 86 130 18
35* 30* 3* 90 110* 14
0.045 0.001 0.001 0.009 0.001 0.001
Statistical significance shown in table is for Mann–Whitney U test between survivors and those who died in each group. ISS, Injury Severity Score; GCS, Glasgow Coma Score; and BP, blood pressure. * Statistically significant difference between older and younger patients who died (Mann–Whitney U test, p < 0.001).
Table 2 Comparison of inpatient mortality rates between groups of patients with various characteristics. Admission variable
Older group (65 and older, 438 patients)
Younger group (under 65, 2229 patients)
Mortality (%)
Risk ratio of death
Significance
Mortality (%)
Risk ratio of death
Significance
100.0 40.7
2.46
0.001
98.9 17.2
5.76
0.001
ISS > 25 Yes No
65.5 31.4
2.08
0.001
34.4 11.6
2.97
0.001
Systolic BP < 90 (mmHg) Yes No
76.6 38.0
2.02
0.001
65.3 15.0
4.35
0.001
Pulse <40 or >90 (bpm) Yes No
50.3 35.9
1.40
0.003
26.0 15.0
1.73
0.002
GCS < 9 Yes No
66.1 31.0
2.13
0.001
87.8 12.2
8.24
0.001
Severe head injury Yes No
47.6 32.4
1.47
0.002
25.3 12.5
2.03
0.001
Severe chest injury Yes No
50.4 39.1
1.29
0.019
28.4 16.6
1.71
0.001
Severe abdominal injury Yes No
61.3 41.0
1.49
0.023
32.3 19.0
1.7
0.001
Cardiac arrest Yes No
Statistical significance assessed using Pearson chi-squared test between survivors and those who died in each group. Risk ratio of inpatient mortality in that age group associated with presence of that variable on admission. Severe is defined as AIS severity score 3 or more recorded for that region. ISS, Injury Severity Score; and BP, blood pressure; GCS, Glasgow Coma Score.
admitted to our unit with severe traumatic injures, and almost 2% of our trauma admissions overall, it is clear they account for a significant part of our practice. In keeping with previous findings, the peak age group in this series was the second decade of life, with the frequency steadily declining after this. As the patients became older, significantly higher proportions were female. There are two likely explanations for this observed change in the male to female ratio with age. Firstly, it is well recognised that high risk behaviour is at a peak in young males.6,11,17,20,23 As this behaviour decreases, the proportion of victims of serious trauma who are male will
likewise decrease. Secondly, higher rates of survival in aging females mean that the proportion of female patients suffering any condition is likely to be higher in older patients.12 Interestingly, logistic regression analysis found male sex to be predictive of survival in the younger patients, the reason for this is unclear. Injury mechanism and distribution Differences between the groups in the causative injury mechanism probably represent different behaviours in older
366
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
Table 3 Models predicting death during initial hospital admission in younger (under 65 years) and older (65 and older) patients.
Model for younger group (2229 patients) Age Sex (male) ISS GCS Cardiac arrest Systolic BP < 90 mmHg Spinal injury Model for older group (438 patients) Age ISS GCS
Odds ratio of mortality
95% Confidence interval for odds ratio
Significance
1.02 0.68 1.08 0.82 49.74 3.12 0.58
1.01–1.03 0.48–0.96 1.06–1.09 0.80–0.85 6.47–382.39 1.96–4.97 0.37–0.91
<0.001 0.031 <0.001 <0.001 <0.001 <0.001 0.017
1.01–1.08 1.06–1.13 0.84–0.92
0.008 <0.001 <0.001
1.04 1.10 0.88
Odds ratio for continuous variables represents the change in the odds of inpatient mortality with each point increase in the variable. Odds ratio for dichotomous variable is the odds ratio of inpatient mortality associated with the presence of that variable on admission. ISS, Injury Severity Score; GCS, Glasgow Coma Score; and BP, blood pressure.
and younger patients. High energy injuries (road traffic accidents and falls from height) continue to be the most common cause of serious trauma in older patients. However, the fact that relatively minor falls were frequently the cause of multiple injuries shows that older people are at increased risk of serious injury following lower energy trauma. This became even more relevant as increasingly older patients were considered. A high index of suspicion must be adopted in searching for concomitant injury in older patients with seemingly innocuous trauma mechanisms. The overall distribution of injury was found to be similar between the groups, though younger patients suffered more trunk injuries compared with older patients. This may be related to differences in the injury mechanisms. Very elderly patients were much more likely to be the victims of severe burns, perhaps related to failures in their social arrangements.1 Length of stay, ICU admission, complications, mortality and age The differences between younger and older patients in terms of complication rates are expected, given changes in physiological reserve with age. Increased rates of mortality in association with these complications in older patients are expected for similar reasons, as well as the overall increase in inpatient mortality with age. In the logistic regression models, age was a significant predictor of mortality for both the younger and older groups. However, even in the oldest patients, survival to discharge rates of over 50% were achieved, confirming that patients should not be denied active treatment on the basis of age alone. It is interesting to note that if only patients over the age of 75 were considered, age was no longer significantly associated with mortality, suggesting that, all other things being equal, a 75-year-old should be treated in the same manner as a 95-year-old regarding potentially lifesaving interventions. Fewer older patients underwent operative treatment for their injuries and fewer were admitted to intensive care units. This is somewhat worrying given older patients’ diminished physiological reserve, increased co-morbidity and increased risk of complications and mortality. This may represent a perceived futility of treatment in older trauma victims, or a less aggressive approach to non-life threatening injuries in this group, which may well be appropriate, particularly in patients with limited preinjury quality of life. However, the fact that overall and ICU length of stay was equivalent between groups suggests that fears over excessively prolonged and expensive treatment courses in older patients are unfounded. Similarly, though there was a trend for older patients to survive longer before ultimately succumbing to their injuries, these differences were not statistically significant for those dying latest after admission.
Outcome prediction The higher mortality rate in the older compared with the younger patients, despite equivalent overall injury severity, suggests that a factor intrinsic to this group increases their risk of death. This increased mortality is in keeping with findings from previous comparative studies.2,8,10 Furthermore, increased mortality in the face of equivalent overall injury severity suggests that the ISS may be less help in predicting which patients might be at increased risk of adverse outcome in the older group. However, logistic regression analysis found the ISS to be a significant predictor of mortality in both groups, and indeed in the over 75’s, with each point increase in the score leading to an increase in the adjusted odds ratio of mortality of 1.076, 1.095 and 1.070 for each group, respectively (p < 0.001 for mortality prediction but no significant difference between the groups). In contrast, various studies have previously found injury scoring systems to be less reliable predictors of outcome in older people.9,10,14 Oreskovich et al., in a series of polytrauma patients older than 65 years of age, found no difference in the ISS between survivors and nonsurvivors.22 Patients who died were significantly more likely to have suffered head injuries or severe burns compared with those who survived. DeMaria et al., in a series of 82 older patients, found that though the mean ISS was higher in non-survivors and correlated with outcome, it was a relatively poor predictor of survival.9 Compared with survivors, non-survivors were older and had a higher rate of medical complications, including cardiac events (54% vs. 10%) and pneumonia (36% vs. 16%). Additionally, non-survivors in both groups had more severe head and neck injuries but there were no differences in the rates of injury to other regions. Horst et al. compared the Trauma Score, ISS and an acute physiology score of 39 patients with multiple injuries, older than 60 years of age, finding no significant differences between survivor and non-survivors.14 These contrasting findings with the current study may in part be due to the relatively small sample size in the previous series. Univariate analysis of admission physiological variables and injury distribution (Tables 1 and 2) grouped by those who died suggests it is more difficult to identify older patients at increased risk. For example, the median admission GCS in the younger patients who died was three compared with nine in the older group. This value was 14 for the survivors in both groups. This is perhaps clearer examining the dichotomous data in Table 2. Cardiac arrest in the older patient carries a 2.46 relative risk of death compared with 5.76 in the adults; a GCS < 9 a relative risk of death of 2.13 in the older patient, compared with 8.24 in the younger adults and a systolic BP < 90 a relative risk of 2.02 in the
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
older patient compared with 4.35 in the younger adult group. Similarly head, chest and abdominal injury (both any injury and severe (AIS 3 or more)) carried a higher relative risk of death in the younger than that in the older group though the differences were less pronounced. This means, for example, that an adult patient with a GCS < 9 is approximately eight times more likely to die as a result of his injuries compared to patients in his age group with a GCS > 9 whereas in the older group, a patient with GCS < 9 is only twice as likely to die compared with an older patient whose GCS is >9. This is universally because the older patient without the finding in question is more likely to die regardless, rather than the younger patient with the abnormality being at increased risk. This finding is confirmed by the results of logistic regression analysis, where only the age, ISS and GCS were independent predictors of mortality in the older patient compared with age, sex, ISS, GCS, cardiac arrest and a systolic BP < 90 on admission. This is again likely to be because the older patients who do not have these characteristics are still at a relatively higher risk of dying due to other factors not considered in the model. Unfortunately data relating to co-morbid conditions were not available in the database for analysis, as this would seem a likely cause for the observed discrepancies and would be in keeping with previous findings.9 This again suggests that a high level of vigilance must be maintained in caring for older victims of major trauma, even in those who do not have the usual indicators of adverse outcome. A lower threshold should be used for transferring these patients to high dependency or intensive care units and serious thought given to early invasive hemodynamic and cardiac monitoring in order to identify occult shock, limit end-organ hypo-perfusion, help prevent multi-organ failure, and ultimately improve survival. Study limitations In applying the findings of this study to clinical practice, it is important that its specific weaknesses are considered. Though the data collation and analysis has been undertaken by a single individual, the data was originally collected by many different people, albeit using standardised forms and coding. This type of data is very prone to individual variation and clinical interpretation bias which is difficult to avoid. As a database analysis, the data available were not specifically collected to answer the questions posed in this study, meaning that various potential confounding parameters were not considered. For example, co-morbidities, poly-pharmacy and poor pre-injury functional status are important to consider, are more likely in the older patients and are not included in this database. In examining the treatment and outcome of a large set of patients it is only possible to determine general trends. The results are not particularly helpful in the management of individual patients, but the findings may prove helpful in tailoring an approach to the care of the older trauma patient in general. Conclusions Older patients accounted for 14% of severely injured patients admitted to our unit and those with multiple injuries for almost 2% of our trauma admissions overall. Though high energy mechanisms were responsible for the majority of these injuries, relatively minor trauma became increasingly important in older patients. Mortality rates in older people were more than twice those seen in the adult
367
population. Age, ISS and Glasgow Coma Score continued to be predictive of mortality in older patients but other factors relevant in younger adults were not. An absence of physiological abnormality in older patients was still associated with a relatively high risk of death, suggesting that it might be more difficult to predict which patients should be treated more aggressively. Despite an increased mortality in older patients, significant survival rates were achieved even in the oldest patients. Active treatment should, therefore, not be withdrawn on the basis of age alone. Conflict of interest The authors are aware of no conflicts of interest, financial or otherwise, related to this work. References 1. Alden NE, Rabbitts A, Yurt RW. Burn injury in patients with dementia: an impetus for prevention. J Burn Care Rehabil 2005;26:267–71. 2. Aldrian S, Nau T, Koenig F, Vecsei V. Geriatric polytrauma. Wien Klin Wochenschr 2005;117:145–9. 3. Baker SP, Harvey AH. Fall injuries in the elderly. Clin Geriatr Med 1985;1:501– 12. 4. Baker SP, O’Neill B, Haddon Jr W, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187–96. 5. Bickel H, Gradinger R, Kochs E, et al. Incidence and risk factors of delirium after hip surgery. Psychiatr Prax 2004;31:360–5. 6. Bina M, Graziano F, Bonino S. Risky driving and lifestyles in adolescence. Accid Anal Prev December 19, 2005. 7. Bitsch M, Foss N, Kristensen B, Kehlet H. Pathogenesis of and management strategies for postoperative delirium after hip fracture: a review. Acta Orthop Scand 2004;75:378–89. 8. Champion HR, Copes WS, Buyer D, et al. Major trauma in geriatric patients. Am J Public Health 1989;79:1278–82. 9. DeMaria EJ, Kenney PR, Merriam MA, et al. Survival after trauma in geriatric patients. Ann Surg 1987;206:738–43. 10. Finelli FC, Jonsson J, Champion HR, et al. A case control study for major trauma in geriatric patients. J Trauma 1989;29:541–8. 11. Green M, Turner C, Purdie DM, McClure R. Injury related risk behavior—a study of Australian skydivers. J Sci Med Sport 2003;6:166–75. 12. Guralnik JM, Balfour JL, Volpato S. The ratio of older women to men: historical perspectives and cross-national comparisons. Aging (Milano) 2000;12:65–76. 13. Hogue CC. Injury in late life. Part I. Epidemiology. J Am Geriatr Soc 1982;30: 183–90. 14. Horst HM, Obeid FN, Sorensen VJ, Bivins BA. Factors influencing survival of elderly trauma patients. Crit Care Med 1986;14:681–4. 15. Kallin K, Jensen J, Olsson LL, et al. Why the elderly fall in residential care facilities, and suggested remedies. J Fam Pract 2004;53:41–52. 16. Krueger PD, Brazil K, Lohfeld LH. Risk factors for falls and injuries in a long-term care facility in Ontario. Can J Public Health 2001;92:117–20. 17. Lam LT. A neglected risky behavior among children and adolescents: underage driving and injury in New South Wales, Australia. J Saf Res 2003;34: 315–20. 18. LEAD-Tools. SPSS. 12.0.1 ed: LEAD Technologies, Inc. 2003. 19. Lecky F, Woodford M, Yates DW. Trends in trauma care in England and Wales 1989–97. UK Trauma Audit and Research Network. Lancet 2000;355: 1771–5. 20. O’Jile JR, Ryan LM, Parks-Levy J, et al. Sensation seeking and risk behaviors in young adults with and without a history of head injury. Appl Neuropsychol 2004;11:107–12. 21. O’Neill TW, Roy DK. How many people develop fractures with what outcome? Best Pract Res Clin Rheumatol 2005;19:879–95. 22. Oreskovich MR, Howard JD, Copass MK, Carrico CJ. Geriatric trauma: injury patterns and outcome. J Trauma 1984;24:565–72. 23. Redeker NS, Smeltzer SC, Kirkpatrick J, Parchment S. Risk factors of adolescent and young adult trauma victims. Am J Crit Care 1995;4:370–8. 24. Shortt NL, Robinson CM. Mortality after low-energy fractures in patients aged at least 45 years old. J Orthop Trauma 2005;19:396–400. 25. Uden G. Inpatient accidents in hospitals. J Am Geriatr Soc 1985;33:833–41. 26. www.statistics.gov.uk. Mid-year population estimates. Office for National Statistics.
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Severe and multiple trauma in older patients; incidence and mortality P.V. Giannoudis a,*, P.J. Harwood a, C. Court-Brown b, H.C. Pape c a
Academic Department Orthopaedic Trauma Surgery, Leeds University, Great George Street, Leeds, West Yorkshire LS1 3EX, UK Edinburgh Orthopaedic Trauma Unit, Royal Infirmary of Edinburgh, Edinburgh, UK c Department of Orthopaedics, University of Pittsburgh Medical Centre, Pittsburgh, USA b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 13 October 2008
Objective: To examine the differences between severely injured older patients (aged over 65 years) compared with similarly injured younger adults in terms of incidence, inpatient mortality and factors predicting outcome. Methods: Data prospectively entered into the Trauma Audit and Research Network (TARN) database from our level I trauma unit over a 5-year period were retrospectively examined, with 3172 patients included in the final analysis. Results: Older patients accounted for 13.8% of those with severe injuries (Injury Severity Score 16 or more) and almost 2% of our trauma admissions overall. High energy injuries were responsible for the majority of these injuries though relatively minor trauma became increasingly important in older patients. Mortality rates in the older patients were more than twice those seen in the adult population (19% in the under 40’s to almost 50% in the over 75’s). Age, Injury Severity Score and Glasgow Coma Score continued to be predictive of mortality in older patients but other factors relevant in younger adults were not. Conclusions: Patients in the older group without physiological derangement on admission were still at a relatively high risk of inpatient mortality. This was in contrast to the younger patients, suggesting that it might be more difficult to predict which older patients might benefit from more aggressive monitoring or treatment. Despite increased mortality in older patients, significant survival rates were achieved even in the oldest. Active treatment should not be withdrawn on the basis of age alone. ß 2008 Elsevier Ltd. All rights reserved.
Keywords: Polytrauma Multiple injury Outcome Mortality Adult Elderly Age
Introduction Background/importance Recent reports of polytrauma in older adults are limited and the subject has previously been given only cursory attention. With an aging, increasingly active older population it is likely that such patients will be seen with increasing frequency.26 Traditional trauma protocols have been established and proven in the treatment of younger accident victims. Older people with diminished physiological reserve, often in association with significant co-morbidities, require special consideration. The distribution of injuries and type of injury mechanism is likely to be different in a population with a high incidence of osteoporosis.21 Older patients can become multiply injured following low energy trauma and these injuries may have worse outcomes.24 Whilst
low energy falls have been reported to account for only 9–11% of injury related deaths in the general population, they comprise more than 50% of traumatic deaths in persons over 65 years old.3,13 Patients with limited mental or physical capacity are also more likely to be involved in accidents as they are slower to identify and respond to dangerous situations.15,16,25 One must also consider the likelihood of a medical emergency such as a myocardial infarction or stroke precipitating an accident, making it necessary to treat this pathology alongside the patient’s injuries. Mental and physical incapacity can make assessment of these patients troublesome. Older patients may become confused and uncooperative during comparatively minor physiological derangements, often compromising their investigation or treatment.5,7 For these reasons it is important to accurately document the pattern of injuries and clinical course of older patients with severe injuries. Goals of the study
* Corresponding author. Tel.: +44 113 2432799. E-mail addresses: [email protected] (P.V. Giannoudis), [email protected] (P.J. Harwood). 0020–1383/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2008.10.016
The aim of this study was to identify and characterise older patients with multiple injuries in our trauma population and
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
compare them to younger patients in terms of their injury patterns, mortality, complications and factors predictive of poor outcome. Materials and methods Data source/study design Data regarding trauma patient care and outcome in our institution is prospectively entered into the Trauma Audit and Research Network (TARN) database. This is an ongoing national study of trauma epidemiology in the United Kingdom.19 The database includes all injured patients arriving alive at the hospital who are then admitted for more than 72 h, or who die in hospital during this initial 72 h period. Patients aged greater than 65 years with isolated fractures of the proximal femur or pubic ramus and patients with simple isolated injuries are excluded. The database includes details of patient demographics, on scene and admission physiology (including pulse, systolic blood pressure, respiratory rate and Glasgow Coma Score), initial resuscitation, inter-hospital transfers and operative intervention. Abbreviated Injury Scale (AIS) injury codes, and the Injury Severity Scores (ISS),4 duration of intensive care unit stay, length of hospital stay or time of death are also documented. This database formed the basis of the study, which therefore represents retrospective analysis of prospectively assembled data. Setting The study included patients admitted to a level one trauma unit, in an urban setting in the United Kingdom. The hospital receives over 7500 admissions per year, directly serves a population of 3.5 million and has access to all medical and surgical specialities on site and acts as a regional tertiary referral centre.
363
Statistical analysis SPSS software was used for statistical analysis (Chicago, IL).18 Parametric and non-parametric tests were used as appropriate to assess continuous variables for significant differences between groups. The Student’s t-test and Mann–Whitney U tests were employed. Dichotomous variables were assessed using a chisquare test. Significant variables were entered into stepwise, multiple logistic regression analysis to determine factors related to outcome. For all outcomes statistical significance was assumed at the p < 0.05 level. The primary outcome used in the study was inpatient mortality from any cause at initial admission. Secondary outcomes included length of inpatient stay, intensive care unit admission, timing of mortality and systemic complications. Results Characteristics of the study group 24,560 patients were entered into the TARN database over the 5-year study period, 13% (3172) were severely injured (ISS equal to or greater than 16). Within the severely injured group, 14% (438) were 65 or older and 16% (505) were under 16. Older patients with severe injuries accounted for 1.8% of the unit’s trauma admissions eligible for entry into the TARN database overall. Fig. 1 shows the frequency distribution of included patients by age. The median age in the younger adult group was 29 (range 16–64 years) compared with 75 (range 65–100 years) in the older group. A significantly greater proportion of patients in the older group were female (42% vs. 22%, p < 0.001). The proportion of female patients remained fairly constant at approximately 22% up to the age 70. Thereafter the proportion of female patients increased dramatically to 41% in those aged 70–80 and 59% in the over 80’s. These increases were statistically significant (p < 0.001).
Selection of patients Analysis of injury severity, hospital stay and treatment Patients admitted to our unit between 1996 and 2001, aged 16 or more, with severe injuries as defined as ISS greater than or equal to 16,4 were identified from the database for inclusion in the study. From this otherwise unselected series, patients were divided into a younger (under 65) and an older (65 or older) group based upon their age at presentation. Further analysis was undertaken dividing the older patient group into 10-year intervals and younger patients into those aged from 16 to 40 and 40 to 65. Data collection and processing Data regarding the above patients were made available from the central database for analysis. The following variables were collected: Demographics and mechanism of injury. Admission physiological variables and details of initial resuscitation. Operative interventions (patients taken to the operating theatre for any treatment). Injury distribution and overall severity (AIS injury codes, and ISS). Duration of intensive care unit admission, hospital stay and time of death. Systemic complications suffered during inpatient treatment including adult respiratory distress syndrome (ARDS), systemic sepsis and multiple organ failure (MOF).
There was no difference in the median ISS between the older and younger adult groups, which was 25 for each. Median length of hospital stay was equivalent at 8 days. If only survivors were considered, the median length of stay was 11 in the younger group and 13 in the older group; this difference was not statistically significant. More patients in the younger group were admitted to ICU and the median length of ICU stay was 4 days (range 0–66) in both groups. 62% of the adult patients underwent operative management during their stay compared with 41% of the older patients (p < 0.001). The proportion of patients receiving operative treatment decreased consecutively with increasing age, from 65% in the under 40’s to 47% in the 65–75’s, 38% in the 75–85’s and 27%
Fig. 1. Age distribution of adult patients admitted between 1996 and 2001 with severe injuries (n = 3172).
364
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
in the over 85’s. In the over 65’s of those operated upon in the first 24 h, 47% underwent neurosurgical, 34% extremity or spinal, 15% abdominal and 7% cardiothoracic interventions. This was not significantly different from the pattern of interventions undertaken in the adult group. No patients over the age of 90 were operated upon. Analysis of injury mechanism and distribution Road traffic accidents (RTA) were the most common cause of injury in each group, though more common in the younger patients (53% vs. 42%, p < 0.001). Falls from height (>2 m) were equally common in each group, whilst minor falls (<2 m) were significantly more common in the older patients (8% vs. 31%, p < 0.001). Younger patients were more likely to be the victims of blunt assault (p < 0.001) or sporting injuries (p = 0.003), both of which were uncommon in either group. Overall anatomic injury distribution was similar between the two groups, 63% of patients had serious (AIS 3 or worse) head, 18% serious thoracic or lower limb injuries and 9% serious abdominal injuries. Younger patients suffered more facial, neck and abdominal injuries overall and more severe (AIS 3 or worse) facial, chest and abdominal injuries when compared with the older patients. Older patients suffered more severe (AIS 3 or worse) external injuries, all of which were severe burns. Analysis of mortality and complications The overall mortality rate was significantly higher in the older patient group (42% vs. 20%, p < 0.001). Fig. 2 shows inpatient mortality by age group. This increased significantly with age from 19% in the under 40’s to almost 50% in the over 75’s. Fig. 3 shows the percentage of those who died in each time period by age group. In both groups, the majority died within the first 24 h, falling progressively with time. However, the overall trend was for a greater proportion of the older patients to die later when compared to the adults. This difference was only statistically significant for the 24–72 h and 2–3-week periods (p < 0.05 and p < 0.001), statistical significance was approached for the two intervening points (p = 0.1). ARDS and MOF were more common in the older than in younger patients (28% vs. 24%, p < 0.05 and 30% vs. 22%, p < 0.001, respectively) whereas systemic sepsis was recorded more commonly in the younger patients (17% vs. 10%, p < 0.01). Considering only the older group, those who died were significantly more likely to have suffered these complications except for sepsis which was not different between the groups (ARDS 34% vs. 23%, p < 0.0001, MOF 35% vs. 14%). Older patients suffering any of these complications were significantly more likely to die compared
Fig. 3. Percentage of deaths occurring in each time period.
with adult patients suffering the same complications (ARDS 51% vs. 23%, sepsis 44% vs. 20%, MOF 59% vs. 32%, all p < 0.0001). Analysis of predictors of outcome on admission Tables 1 and 2 compare various admission variables between survivors and those who died in each group. In the older group, survivors were significantly younger, had a significantly lower ISS, a lower pulse and higher GCS and BP on admission. In the adult group, as well as these parameters, survivors received significantly less fluid in the accident and emergency department and had significantly higher respiratory rates compared with those who died. Cardiac arrest, GCS < 9, BP < 90 on admission and ISS > 25 were significantly associated with increased mortality in both groups. A pulse greater than 90 was associated with increased mortality in the older patients group but not the younger adults. Any head, chest or abdominal injury as well as severe (AIS severity 3 or greater) head, chest, abdominal and spinal injuries were associated with a significantly increased rate of mortality in both groups. Any spinal or external injury was associated with decreased risk of mortality in the adult but not the older group, whilst the risk associated with severe lower limb injuries approached statistical significance in the older but not the younger patients. Other factors considered were not associated with any difference in mortality rates. Table 3 summarises the results of logistic regression analysis carried out on data from the two groups. Variables under consideration were ISS, age, sex, head, chest, abdominal or spinal injuries, cardiac arrest, and admission blood pressure, pulse and respiratory rate, with these physiological values considered both as raw continuous and dichotomous variables (pulse > 90, systolic BP < 90, respiratory rate > 20). For the adult patients, age, female sex, ISS, Glasgow Coma Score, cardiac arrest on admission and systolic blood pressure < 90 were predict as of mortality. The presence of any spinal injury predicted survival (adjusted odds ratio 0.577). For the older patients, only age, ISS and Glasgow Coma Score were significant predictors of mortality. If only patients over the age of 75 are considered, only the ISS and Glasgow Coma Score were significant variables included in the model (adjusted odds ratios 1.07 and 0.83 change per point increase in each score, respectively, both independent predictors of mortality p < 0.001). Discussion Incidence and demographics
Fig. 2. Percentage mortality by age group. *Significant difference between consecutive groups, p < 0.05.
As the population ages and older people become increasingly healthy and active, older victims of multiple trauma will become more frequent. With older people making up 14% of the patients
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
365
Table 1 Characteristics of survivors and those who died during initial hospital admission. Admission variable (median)
Age (years) ISS (16–75) GCS (3–15) Pulse (bpm) Systolic BP (mmHg) Respiratory rate (per minute)
Older group (65 and over, 438 patients)
Younger group (Under 65, 2229 patients)
Survived (254 patients)
Died (184 patients)
Significance
Survived (1783 patients)
Died (446 patients)
Significance
74 20 14 80 140 18
76* 26* 9* 88 130* 18
0.013 0.001 0.001 0.021 0.024 Not significant
33 25 14 86 130 18
35* 30* 3* 90 110* 14
0.045 0.001 0.001 0.009 0.001 0.001
Statistical significance shown in table is for Mann–Whitney U test between survivors and those who died in each group. ISS, Injury Severity Score; GCS, Glasgow Coma Score; and BP, blood pressure. * Statistically significant difference between older and younger patients who died (Mann–Whitney U test, p < 0.001).
Table 2 Comparison of inpatient mortality rates between groups of patients with various characteristics. Admission variable
Older group (65 and older, 438 patients)
Younger group (under 65, 2229 patients)
Mortality (%)
Risk ratio of death
Significance
Mortality (%)
Risk ratio of death
Significance
100.0 40.7
2.46
0.001
98.9 17.2
5.76
0.001
ISS > 25 Yes No
65.5 31.4
2.08
0.001
34.4 11.6
2.97
0.001
Systolic BP < 90 (mmHg) Yes No
76.6 38.0
2.02
0.001
65.3 15.0
4.35
0.001
Pulse <40 or >90 (bpm) Yes No
50.3 35.9
1.40
0.003
26.0 15.0
1.73
0.002
GCS < 9 Yes No
66.1 31.0
2.13
0.001
87.8 12.2
8.24
0.001
Severe head injury Yes No
47.6 32.4
1.47
0.002
25.3 12.5
2.03
0.001
Severe chest injury Yes No
50.4 39.1
1.29
0.019
28.4 16.6
1.71
0.001
Severe abdominal injury Yes No
61.3 41.0
1.49
0.023
32.3 19.0
1.7
0.001
Cardiac arrest Yes No
Statistical significance assessed using Pearson chi-squared test between survivors and those who died in each group. Risk ratio of inpatient mortality in that age group associated with presence of that variable on admission. Severe is defined as AIS severity score 3 or more recorded for that region. ISS, Injury Severity Score; and BP, blood pressure; GCS, Glasgow Coma Score.
admitted to our unit with severe traumatic injures, and almost 2% of our trauma admissions overall, it is clear they account for a significant part of our practice. In keeping with previous findings, the peak age group in this series was the second decade of life, with the frequency steadily declining after this. As the patients became older, significantly higher proportions were female. There are two likely explanations for this observed change in the male to female ratio with age. Firstly, it is well recognised that high risk behaviour is at a peak in young males.6,11,17,20,23 As this behaviour decreases, the proportion of victims of serious trauma who are male will
likewise decrease. Secondly, higher rates of survival in aging females mean that the proportion of female patients suffering any condition is likely to be higher in older patients.12 Interestingly, logistic regression analysis found male sex to be predictive of survival in the younger patients, the reason for this is unclear. Injury mechanism and distribution Differences between the groups in the causative injury mechanism probably represent different behaviours in older
366
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
Table 3 Models predicting death during initial hospital admission in younger (under 65 years) and older (65 and older) patients.
Model for younger group (2229 patients) Age Sex (male) ISS GCS Cardiac arrest Systolic BP < 90 mmHg Spinal injury Model for older group (438 patients) Age ISS GCS
Odds ratio of mortality
95% Confidence interval for odds ratio
Significance
1.02 0.68 1.08 0.82 49.74 3.12 0.58
1.01–1.03 0.48–0.96 1.06–1.09 0.80–0.85 6.47–382.39 1.96–4.97 0.37–0.91
<0.001 0.031 <0.001 <0.001 <0.001 <0.001 0.017
1.01–1.08 1.06–1.13 0.84–0.92
0.008 <0.001 <0.001
1.04 1.10 0.88
Odds ratio for continuous variables represents the change in the odds of inpatient mortality with each point increase in the variable. Odds ratio for dichotomous variable is the odds ratio of inpatient mortality associated with the presence of that variable on admission. ISS, Injury Severity Score; GCS, Glasgow Coma Score; and BP, blood pressure.
and younger patients. High energy injuries (road traffic accidents and falls from height) continue to be the most common cause of serious trauma in older patients. However, the fact that relatively minor falls were frequently the cause of multiple injuries shows that older people are at increased risk of serious injury following lower energy trauma. This became even more relevant as increasingly older patients were considered. A high index of suspicion must be adopted in searching for concomitant injury in older patients with seemingly innocuous trauma mechanisms. The overall distribution of injury was found to be similar between the groups, though younger patients suffered more trunk injuries compared with older patients. This may be related to differences in the injury mechanisms. Very elderly patients were much more likely to be the victims of severe burns, perhaps related to failures in their social arrangements.1 Length of stay, ICU admission, complications, mortality and age The differences between younger and older patients in terms of complication rates are expected, given changes in physiological reserve with age. Increased rates of mortality in association with these complications in older patients are expected for similar reasons, as well as the overall increase in inpatient mortality with age. In the logistic regression models, age was a significant predictor of mortality for both the younger and older groups. However, even in the oldest patients, survival to discharge rates of over 50% were achieved, confirming that patients should not be denied active treatment on the basis of age alone. It is interesting to note that if only patients over the age of 75 were considered, age was no longer significantly associated with mortality, suggesting that, all other things being equal, a 75-year-old should be treated in the same manner as a 95-year-old regarding potentially lifesaving interventions. Fewer older patients underwent operative treatment for their injuries and fewer were admitted to intensive care units. This is somewhat worrying given older patients’ diminished physiological reserve, increased co-morbidity and increased risk of complications and mortality. This may represent a perceived futility of treatment in older trauma victims, or a less aggressive approach to non-life threatening injuries in this group, which may well be appropriate, particularly in patients with limited preinjury quality of life. However, the fact that overall and ICU length of stay was equivalent between groups suggests that fears over excessively prolonged and expensive treatment courses in older patients are unfounded. Similarly, though there was a trend for older patients to survive longer before ultimately succumbing to their injuries, these differences were not statistically significant for those dying latest after admission.
Outcome prediction The higher mortality rate in the older compared with the younger patients, despite equivalent overall injury severity, suggests that a factor intrinsic to this group increases their risk of death. This increased mortality is in keeping with findings from previous comparative studies.2,8,10 Furthermore, increased mortality in the face of equivalent overall injury severity suggests that the ISS may be less help in predicting which patients might be at increased risk of adverse outcome in the older group. However, logistic regression analysis found the ISS to be a significant predictor of mortality in both groups, and indeed in the over 75’s, with each point increase in the score leading to an increase in the adjusted odds ratio of mortality of 1.076, 1.095 and 1.070 for each group, respectively (p < 0.001 for mortality prediction but no significant difference between the groups). In contrast, various studies have previously found injury scoring systems to be less reliable predictors of outcome in older people.9,10,14 Oreskovich et al., in a series of polytrauma patients older than 65 years of age, found no difference in the ISS between survivors and nonsurvivors.22 Patients who died were significantly more likely to have suffered head injuries or severe burns compared with those who survived. DeMaria et al., in a series of 82 older patients, found that though the mean ISS was higher in non-survivors and correlated with outcome, it was a relatively poor predictor of survival.9 Compared with survivors, non-survivors were older and had a higher rate of medical complications, including cardiac events (54% vs. 10%) and pneumonia (36% vs. 16%). Additionally, non-survivors in both groups had more severe head and neck injuries but there were no differences in the rates of injury to other regions. Horst et al. compared the Trauma Score, ISS and an acute physiology score of 39 patients with multiple injuries, older than 60 years of age, finding no significant differences between survivor and non-survivors.14 These contrasting findings with the current study may in part be due to the relatively small sample size in the previous series. Univariate analysis of admission physiological variables and injury distribution (Tables 1 and 2) grouped by those who died suggests it is more difficult to identify older patients at increased risk. For example, the median admission GCS in the younger patients who died was three compared with nine in the older group. This value was 14 for the survivors in both groups. This is perhaps clearer examining the dichotomous data in Table 2. Cardiac arrest in the older patient carries a 2.46 relative risk of death compared with 5.76 in the adults; a GCS < 9 a relative risk of death of 2.13 in the older patient, compared with 8.24 in the younger adults and a systolic BP < 90 a relative risk of 2.02 in the
P.V. Giannoudis et al. / Injury, Int. J. Care Injured 40 (2009) 362–367
older patient compared with 4.35 in the younger adult group. Similarly head, chest and abdominal injury (both any injury and severe (AIS 3 or more)) carried a higher relative risk of death in the younger than that in the older group though the differences were less pronounced. This means, for example, that an adult patient with a GCS < 9 is approximately eight times more likely to die as a result of his injuries compared to patients in his age group with a GCS > 9 whereas in the older group, a patient with GCS < 9 is only twice as likely to die compared with an older patient whose GCS is >9. This is universally because the older patient without the finding in question is more likely to die regardless, rather than the younger patient with the abnormality being at increased risk. This finding is confirmed by the results of logistic regression analysis, where only the age, ISS and GCS were independent predictors of mortality in the older patient compared with age, sex, ISS, GCS, cardiac arrest and a systolic BP < 90 on admission. This is again likely to be because the older patients who do not have these characteristics are still at a relatively higher risk of dying due to other factors not considered in the model. Unfortunately data relating to co-morbid conditions were not available in the database for analysis, as this would seem a likely cause for the observed discrepancies and would be in keeping with previous findings.9 This again suggests that a high level of vigilance must be maintained in caring for older victims of major trauma, even in those who do not have the usual indicators of adverse outcome. A lower threshold should be used for transferring these patients to high dependency or intensive care units and serious thought given to early invasive hemodynamic and cardiac monitoring in order to identify occult shock, limit end-organ hypo-perfusion, help prevent multi-organ failure, and ultimately improve survival. Study limitations In applying the findings of this study to clinical practice, it is important that its specific weaknesses are considered. Though the data collation and analysis has been undertaken by a single individual, the data was originally collected by many different people, albeit using standardised forms and coding. This type of data is very prone to individual variation and clinical interpretation bias which is difficult to avoid. As a database analysis, the data available were not specifically collected to answer the questions posed in this study, meaning that various potential confounding parameters were not considered. For example, co-morbidities, poly-pharmacy and poor pre-injury functional status are important to consider, are more likely in the older patients and are not included in this database. In examining the treatment and outcome of a large set of patients it is only possible to determine general trends. The results are not particularly helpful in the management of individual patients, but the findings may prove helpful in tailoring an approach to the care of the older trauma patient in general. Conclusions Older patients accounted for 14% of severely injured patients admitted to our unit and those with multiple injuries for almost 2% of our trauma admissions overall. Though high energy mechanisms were responsible for the majority of these injuries, relatively minor trauma became increasingly important in older patients. Mortality rates in older people were more than twice those seen in the adult
367
population. Age, ISS and Glasgow Coma Score continued to be predictive of mortality in older patients but other factors relevant in younger adults were not. An absence of physiological abnormality in older patients was still associated with a relatively high risk of death, suggesting that it might be more difficult to predict which patients should be treated more aggressively. Despite an increased mortality in older patients, significant survival rates were achieved even in the oldest patients. Active treatment should, therefore, not be withdrawn on the basis of age alone. Conflict of interest The authors are aware of no conflicts of interest, financial or otherwise, related to this work. References 1. Alden NE, Rabbitts A, Yurt RW. Burn injury in patients with dementia: an impetus for prevention. J Burn Care Rehabil 2005;26:267–71. 2. Aldrian S, Nau T, Koenig F, Vecsei V. Geriatric polytrauma. Wien Klin Wochenschr 2005;117:145–9. 3. Baker SP, Harvey AH. Fall injuries in the elderly. Clin Geriatr Med 1985;1:501– 12. 4. Baker SP, O’Neill B, Haddon Jr W, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187–96. 5. Bickel H, Gradinger R, Kochs E, et al. Incidence and risk factors of delirium after hip surgery. Psychiatr Prax 2004;31:360–5. 6. Bina M, Graziano F, Bonino S. Risky driving and lifestyles in adolescence. Accid Anal Prev December 19, 2005. 7. Bitsch M, Foss N, Kristensen B, Kehlet H. Pathogenesis of and management strategies for postoperative delirium after hip fracture: a review. Acta Orthop Scand 2004;75:378–89. 8. Champion HR, Copes WS, Buyer D, et al. Major trauma in geriatric patients. Am J Public Health 1989;79:1278–82. 9. DeMaria EJ, Kenney PR, Merriam MA, et al. Survival after trauma in geriatric patients. Ann Surg 1987;206:738–43. 10. Finelli FC, Jonsson J, Champion HR, et al. A case control study for major trauma in geriatric patients. J Trauma 1989;29:541–8. 11. Green M, Turner C, Purdie DM, McClure R. Injury related risk behavior—a study of Australian skydivers. J Sci Med Sport 2003;6:166–75. 12. Guralnik JM, Balfour JL, Volpato S. The ratio of older women to men: historical perspectives and cross-national comparisons. Aging (Milano) 2000;12:65–76. 13. Hogue CC. Injury in late life. Part I. Epidemiology. J Am Geriatr Soc 1982;30: 183–90. 14. Horst HM, Obeid FN, Sorensen VJ, Bivins BA. Factors influencing survival of elderly trauma patients. Crit Care Med 1986;14:681–4. 15. Kallin K, Jensen J, Olsson LL, et al. Why the elderly fall in residential care facilities, and suggested remedies. J Fam Pract 2004;53:41–52. 16. Krueger PD, Brazil K, Lohfeld LH. Risk factors for falls and injuries in a long-term care facility in Ontario. Can J Public Health 2001;92:117–20. 17. Lam LT. A neglected risky behavior among children and adolescents: underage driving and injury in New South Wales, Australia. J Saf Res 2003;34: 315–20. 18. LEAD-Tools. SPSS. 12.0.1 ed: LEAD Technologies, Inc. 2003. 19. Lecky F, Woodford M, Yates DW. Trends in trauma care in England and Wales 1989–97. UK Trauma Audit and Research Network. Lancet 2000;355: 1771–5. 20. O’Jile JR, Ryan LM, Parks-Levy J, et al. Sensation seeking and risk behaviors in young adults with and without a history of head injury. Appl Neuropsychol 2004;11:107–12. 21. O’Neill TW, Roy DK. How many people develop fractures with what outcome? Best Pract Res Clin Rheumatol 2005;19:879–95. 22. Oreskovich MR, Howard JD, Copass MK, Carrico CJ. Geriatric trauma: injury patterns and outcome. J Trauma 1984;24:565–72. 23. Redeker NS, Smeltzer SC, Kirkpatrick J, Parchment S. Risk factors of adolescent and young adult trauma victims. Am J Crit Care 1995;4:370–8. 24. Shortt NL, Robinson CM. Mortality after low-energy fractures in patients aged at least 45 years old. J Orthop Trauma 2005;19:396–400. 25. Uden G. Inpatient accidents in hospitals. J Am Geriatr Soc 1985;33:833–41. 26. www.statistics.gov.uk. Mid-year population estimates. Office for National Statistics.