Patients with burn injuries admitted to Norwegian hospitals – a population-based study

Material and method

In September 2013, data were retrieved from among all admissions to Norwegian hospitals that provide somatic emergency services. The search included all patients discharged during the period 1 January  –  31 December 2012 with burn injuries (ICD-10) (2): T20  –  T25, T29 or T30, all with .0  –  .3 as the third digit, or T31.0  –  T31.9 as their main diagnosis. Patients with smoke inhalation injuries (T58 or T59.7  –  T59.9) as their main diagnosis and burn injuries as a secondary diagnosis were also included.

This raw material was then processed further before it was supplied from the Norwegian Patient Registry, so as to exclude patients who were only registered with an admission for burn injuries with a hospitalisation period of 0 days (Figure 1). The remaining data included 620 unique patients (with a total of 1 976 admissions) who were subjected to further analysis.

Figure 1  Selection of the study group: admissions to hospitals for burn injuries in 2012 (from the Norwegian Patient Registry)

The anonymised data were supplied by the Norwegian Patient Registry in a tabular format. For individuals, the information included only their year of birth, gender, number of admissions with burn injuries as the main diagnosis and total hospitalisation time. In age groups with few patients, age was stated as intervals (e.g. 78 – 82 years), and for these patients age has been estimated as the mean value of the stated age interval (80 years in the example).

Hospitalisation time has been calculated as the date of discharge minus the date of admission. For patients with multiple admissions, the total hospitalisation time was calculated as the duration of two or more hospitalisation periods.

Age has been calculated as 2012 minus the year of birth, since access to dates of birth was not granted for reasons of privacy protection. This gives rise to a certain degree of systematic imprecision, and the estimated incidence of burn injuries in the youngest age group (< 3 years) may thus be somewhat low.

At the institutional level, information was supplied on the total number of patients, number of admissions and the total number of hospitalisation days.

To elucidate the causes of burn injuries in little children, additional information on the type of injury, circumstances, area of injury, hospitalisation times and number of operations was retrieved from a local quality registry at the Burn Unit, Haukeland University Hospital, Bergen, for all children < 3 years who were admitted during the period 2013 – 2015.

Incidence has been estimated by total population and gender in various age groups in Norway as of 1 January 2012 (total population 4 985 870) (3).

Results

The data retrieved from the Norwegian Patient Registry included 620 individual patients with burn injuries (Figure 1). This is equivalent to 12.4 patients admitted for burn injuries per 100 000 inhabitants per year.

Altogether 393 patients (63.4 %) were men. Figure 2 shows the number of patients with burn injuries per 100 000 inhabitants in various age groups and by gender. There was a significant preponderance of burn injuries among males, in children < 3 years as well as the age groups 10 – 59 years.

Figure 2  Age-adjusted incidence of patients with burn injuries admitted to Norwegian hospitals in 2012. Based on data retrieved from the Norwegian Patient Registry. Information on the number of inhabitants in different age groups was retrieved from Statistics Norway (3). *p < 0.05; **p < 0.005; ***p < 0.0001 for gender differences

More than one-third of the patients with burn injuries were children, and as many as 216 patients (34.8 %) were younger than five years. The large proportion of young children accounts for the low average age of the patients with burn injuries: 27.4 years (SD (26.0 years), with a median age of 22.5 years.

The total number of hospitalisation days (n = 6 981) indicates that on average, 20 Norwegian hospital beds are occupied by patients with burn injuries on a daily basis. The hospitalisation time was strongly skewed towards short hospitalisation periods (average hospitalisation time amounted to 11.3 days (SD 18.8 days), median hospitalisation time was five days) (Table 1). Altogether 367 patients (59.2 %) had been hospitalised for less than eight days. Those 15 patients whose hospitalisation periods exceeded eight weeks were at the other extreme, with a total of 1 438 hospitalisation days, equivalent to 20.6 % of all hospitalisation days due to burn injuries.

The 620 patients were registered with a total of 1 976 admissions. Altogether 375 patients (60 %) had been admitted more than once. Multiple admissions of the same patients are likely to reflect that many patients with burn injuries return for one (or more) change of bandages/wound care, that some are readmitted with persistent wound problems, and that patients with burn injuries are transferred between hospitals. One patient in every five had been admitted for burn injuries to more than one hospital. Of these, altogether 109 patients (17.6 %) had been admitted to two hospitals, and 15 patients (3.4 %) had been admitted to three or more hospitals.

In addition to those 620 patients who had been hospitalised for burn injuries, another 2 154 patients had been in contact with a hospital after sustaining burn injuries, but without being admitted (Figure 1).

Haukeland University Hospital functions as a national centre for advanced treatment of burn injuries. In 2012, a total of 103 of the 620 patients (16.7 %) had been treated in this hospital, with a total of 2 148 hospitalisation days for burn injuries (30.7 % of all hospitalisation days).

Seven patients, three men and four women, were registered as deceased on discharge. Six of them had hospitalisation periods ranging from one day to three weeks, while the last one died after a lengthy period in hospital. Of those whose burn injuries proved fatal, one patient was younger than 15 years and four were 70 years or older.

Altogether 183 of the patients (29.5 %) were younger than three years. In this age group, burn injuries thus occur 12 times more frequently than among adults and children older than five years (Figure 2).

In the period 2013 – 2015, a total of 344 patients with burn injuries were admitted to the Burn Unit at Haukeland University Hospital. Altogether 114 (33 %) of these patients were children < 3 years. Of these, as many as 89 (78 %) had suffered scalding injuries, while 19 (17 %) had sustained burns from contact with hot surfaces (Table 2).

Discussion

Our review of patients with burn injuries admitted to all Norwegian hospitals in 2012 shows that this type of injury remains a substantial cause of hospitalisation. What is most disturbing is the fact that children under three years are admitted 12 times more frequently than children ≥ 5 years and adults. Many burn injuries in young children can be prevented through greater vigilance regarding hazardous situations, especially in the home: hot liquids, unsecured hot-water taps and hot surfaces (wood-burning stoves).

The retrieval of data from the Norwegian Patient Registry provides an opportunity for collating information linked to patients with multiple admissions (readmissions) or multiple stays in different hospitals for the same injury.

This study has methodological limitations: at the individual level, we only had access to information on year of birth, gender, the number of admissions and the total number of hospitalisation days. Access to more detailed information at the individual level, especially about diagnoses and coding for the external cause of injury, might have provided further opportunities for analysis without any real risk of compromising confidentiality.

A strength of this study, however, is that the source data can be assumed to be near-complete. The number of patients identified as having sustained burn injuries will nevertheless be a minimum number. To identify patients with acute burn injuries, without including patients who had burn injuries as a (more or less coincidental) secondary diagnosis, we searched for patients who had burn injuries as their main diagnosis (or with smoke inhalation injuries as their main diagnosis and burn injuries as a secondary diagnosis). Thereby, we may potentially have failed to capture certain patients with burn injuries, for example if A41.0 (sepsis due to Staphylococcus aureus, which is a real complication) was stated as the main diagnosis and the burn injury as a secondary diagnosis. The hospitalisation periods may also be slightly underestimated, since we did not include all stays for the relatively few patients who had multiple admissions before year-end as well as early in the new year.

The total scope of burn injury treatment may also be slightly underestimated, since we may potentially have failed to capture patients who were readmitted with other complications in their course of treatment after a burn injury. Nor did the data retrieval include admissions to further rehabilitation after a burn injury, nor did we focus on those who were hospitalised to undergo reconstructive surgery as part of the treatment sequence after a burn injury.

In addition to our 620 patients, another 2 154 persons had contacted a hospital for burn injuries, but without being admitted (Figure 1). In other words, somewhat less than 3 000 persons with burn injuries contact Norwegian hospitals each year. In addition to these, a considerable number of minor burn injuries are treated in medical centres and local A&E clinics.

Patients with burn injuries represent a very heterogeneous group when it comes to the extent of the damaged skin surface and the depth of injury. Experience from a previous Norwegian study showed that coding for burns according to the extent of body surface involved (ICD-10; T31.0  –  T31.9) had been entered so rarely (in only 23 %) (4) as to provide an insufficient basis for analysis of the extent of the injuries. Nor did the present material include available codes for external causes of diseases, injuries and deaths (V0n-Y98), which might have helped provide information on the causes of injury.

In each of the years 1999 – 2010, an average of 58 people died in fires (3). The vast majority of those who perish do so at the site of the fire. Of the 620 patients who were admitted to hospital with burn injuries in 2012, seven died (1.1 %). This is a lower proportion of deaths than was found in two previous Norwegian studies (1, 4), but in line with a national study of burn injuries undertaken in England in 2010, in which the 30-day mortality of patients with such injuries amounted to 0.9 % (5).

European studies show a great variation in the reported incidence of burn injuries: from two to 29 hospitalised burn victims per 100 000 inhabitants per year (6). For Norway, in this study we have found an incidence of burn injuries requiring hospitalisation equivalent to 12.4 per 100 000 inhabitants per year. Our figures match those from Sweden (for 2004): 13.1 patients with burn injuries per 100 000 inhabitants per year (7), as well as a Dutch study from 2011 reporting 8.5 patients with burn injuries per 100 000 inhabitants per year (8). In a study from England, on the other hand, Brewster and collaborators (5) reported 24 patients with burn injuries per 100 000 inhabitants per year, but this last figure had not been corrected for readmissions.

Altogether 30 % of all hospitalisation days caused by burn injuries in 2012 were associated with stays at Haukeland University Hospital. In the period 2008 – 2015, we noticed an increase in activity at the Burn Unit – there was a 60 % increase in the number of referrals and a 60 % increase in hospitalisation days (Onarheim, unpublished data). Even though the total number of burn victims or the extent of burns is not necessarily increasing, a larger proportion currently seem to be referred to specialist care units. A similar tendency was observed in the Netherlands in the period 1995 – 2010: a 60 % increase in the number of patients treated in the three Dutch burn units, while the number of burn injuries nationwide remained constant (8).

The collation of our incidence data with figures from four earlier Norwegian studies conducted over a period of 35 years (1, 4, 9, 10) showed that the number of burn injuries in 2012 was 35 % lower than in the oldest material, with data from 1977 (10) (Table 3). Because of differences in the procedure used to retrieve data and in the opportunities to correct for readmissions and transfers between hospitals, it is not possible to conclude with any certainty that we have been successful in reducing the incidence of burn injuries in 2012, judging from the number of burn victims admitted to hospital in Norway when compared to the situation twenty years earlier (9).