**3. Results**

### *3.1. Incidence of Hospitalization and In-Hospital Fatality Rate*

During 2007–2008, 476,241 cases were retrieved as trauma admissions. There were 50,599 cases of mild head injury in which the AIS in the head region was 1 or 2. Among these cases, 151 deaths occurred and the in-hospital mortality rate was 0.3%. There were 9666 cases of moderate head injury (AIS head of 3) and 126 deaths were recognized; the in-hospital mortality rate was 1.3%. There were 39,126 cases of severe head injury (AIS head of 4, 5, or 6) and 4809 deaths were recognized; the in-hospital mortality rate was 12.3%. The raw hospitalization rate for all head injuries in Taiwan was 215.3 per 100,000, and for moderate-to-severe head injury, 105.9 per 100,000. After calculations using the U.S. 2000 standard population, the standardized incidence of all head injury was 220.6 per 100,000 in Taiwan and the standardized incidence of moderate-to-severe head injury was 110.5 per 100,000. Table 3 shows the incidence of trauma hospitalization with head injury in the nine age strata. The lowest trauma hospitalization rate for patients with moderate-to-severe head injury in Taiwan was in the 0–14 year age group (29.8/100,000), followed by a higher incidence in the 15–24 year age group (97.9/199,999), and gradually increased up to 582.4/100,000 in the age group older than 84 years. The graph pattern of hospitalization rate in the nine age strata and the in-hospital fatality rate for patients with moderate-to-severe head injury are shown in Figure 1. The in-hospital mortality rate for patients with moderate-to-severe head injury was 10.7% and the standardized hospitalization death rate was 11.8/100,000. The hospitalization death rate for patients with moderate-to-severe head injury increased gradually from 1.2/100,000 in the group younger than 15 years old to the highest rate, 108.5/100,000, in the group older than 84 years, except for slightly lower rates in the 25–34 year age group than in the 15–24 year age group. Men had higher rates than women in all age groups.

**Figure 1.** Incidence rate of head injury admission and death by age groups.


**Table 3.** Age distribution of trauma hospitalizations with head injury in Taiwan in 2007–2008.

In Taiwan, 6,935,205 people resided in rural areas in 2008. During the study period, 41,570 rural people were admitted with head injury and 18,108 of these rural people had moderate-to severe head injury. Thus, focusing on moderate-to-severe head injury, the patients who resided in rural areas had a higher hospitalization rate than those residing in urban areas (130.6 vs. 95.3 per 100,000, respectively; *p* < 0.01). Among the patients who had in-hospital mortality, 2038 were registered in rural areas. The rural population yielded a higher raw hospitalization death rate for patients with moderate-to-severe head injury than the urban population (14.2 vs. 9.2 per 100,000, respectively; *p* = 0.005). Monthly distribution of case numbers during 2007–2008 is shown in Figure 2, and in general, the occurrence did not significantly change with month.

**Figure 2.** Monthly distribution of case numbers during 2007–2008.

Table 4 presents the demographics of two groups: all patients with head injury and those with moderate-to-severe head injury. Table 2 summarizes the comparisons of age, sex, and other demographics, the associated injury characteristics, and distribution of the locations of treatment between these two groups. The patients with moderate-to-severe injury were older (50.6 ± 23.1 vs. 46.8 ± 22.9 years), a higher proportion were men, and rural people had a higher Charlson comorbidity index and a higher ICDISS score (18.2 ± 6.4 vs. 12.0 ± 8.1 years), and had a longer median stay in the hospital (7, 4–13 vs. 5, 2–6 (days, 25th–75th centile). In the moderate-to-severe head injury group, more patients were treated at a medical center and fewer at a local hospital. Figure 3 shows the different distribution of the mechanisms of injury among these nine age groups. Among those in the younger population, being injured as a motorcycle rider or passenger was more frequently the external cause, and a fall on the same level was more frequently the cause in older patients. The proportion of case numbers for motorcycle injury in people aged 15–24 years was significantly higher than that in people of other ages (49.4% vs. 26.4%, *p* < 0.001).



**Figure 3.** Mechanisms of head injury by age groups.

### *3.2. Results of Regression Analyses*

We included characteristics of the patients and their injuries, including age, sex, rural residency, level of monthly income, Charlson comorbidity index, mechanism of injury, and ICD/ISS in a logistic regression to estimate the adjusted relative odds of in-hospital fatality for the patients with moderate-to-severe head injury. The results are shown in Table 5. Different age strata experienced different effects; being younger than 15 years decreases the risk of death, but being older than 54 years is associated with a higher risk of death. The older the patient, the higher the hospitalization mortality rate. Rural residency increases the risk of death (odds ratio (OR): 1.19, 95% confidence interval (CI): 1.11–1.27). Compared with the lowest level of monthly income (less than USD \$660), the group with a monthly income between USD \$660 and 1320 had a lower probability of death (OR: 0.78, 95% CI: 0.72–0.85), but dependents had a higher probability of in-hospital death (OR: 1.49, 95% CI: 1.38–1.61) Compared with the patients treated at medical centers, being treated at a regional hospital was associated with a higher probability of in-hospital fatality (OR: 1.11, 95% CI: 1.03–1.19). Compared with the patients treated at private hospitals, being treated at a public hospital was associated with slightly lower risk of death (OR: 0.92, 95% CI: 0.86–1.00).

As an indicator of PECs, higher Charlson comorbidity index scores were associated with a higher risk of death. The relative odds of dying in the group with a Charlson comorbidity index of two was 1.52 times higher than those for the group with a Charlson comorbidity index of zero (95% CI: 1.34–1.72) and the relative odds of dying in the group with a Charlson comorbidity index more than two was 2.5 times higher (95% CI: 2.16–2.90). Several mechanisms of injury did not significantly affect the probability of death. Compared with falls on the same level, only pedestrians hit by a motor vehicle and falls from a height had a higher probability of in-hospital death. Furthermore, being associated with another significant injury significantly increased the risk of mortality; the mortality rate in the group associated with a significant abdominal, chest, or lower extremity injury was 3.41, 3.41, and 1.37 times higher, respectively, than the group without a significant injury.


