https://rpubs.com/alex_istrate/588829

1 General part

2 Special part

2.1 Introduction

2.1.1 Objectives

2.2 Methods

2.2.1 Statisitcal analysis of the data

To describe the quantitative variables, we used histograms and we calculated the arithmetic means ±standard deviation (SD) [with 95% confidence interval (CI)], as well as the extreme and median values. For the qualitative variables, we used pie or bar charts and calculated the absolute and percentual frequencies of the formed categories.

To study the relationships between quantitative and qualitative variables, we used the T or Mann-Whitney (MW) tests. We presented the p values generated by these tests as well as the means ±SD of the groups and the difference of the means with associated 95% CI. To study the relationships between the quantitative variables we used the Spearman correlation coefficient (R), with the associated p-value and graphically presented the relations as scatter plots, on which we added the regression line. To describe the relationships between the qualitative variables we used the Chi² or Fisher test and the Cramer V and Odds-Ratio (OR) indicators with 95% CI. We graphically presented the results in the form of pie or bar charts.

We used Microsoft Excel 2016 for database management. For all statistical analyzes and subsequent graphs we used R 3.6.3 [1]. We considered p <0.05 to be statistically significant and p <0.10 to show only a tendency towards statistical significance.

2.3 Results

2.3.1 Demographics

A total of 50 patients were included in the study, of whom 75% were men. Age had values between 14 and 89 years old (median = 61) with a mean of 58.36 ±18.8 years old. Men were approximatively 8.69 years younger on average than men (Mann-Whitney test: p=0.054, showing a tendency towards statistical significane). Most cases were from Cluj (77.8%) and some were from neighboring places (Salaj, Baia Mare and Satu Mare counties, Dej, Turda).

Table 1: Demographic parameters of the sample.

Variable

Details

Total

N=

81

Sex

F

20 (24.7%)

M

61 (75.3%)

Age (years)

μ ±SD

58.36 ±18.8

M (min:max)

61 (14:89)

Residence

Cluj

63 (77.8%)

Dej

6 (7.4%)

Salaj

6 (7.4%)

Satu mare

3 (3.7%)

Baia mare

2 (2.5%)

Turda

1 (1.2%)

μ ±SD = Mean (standard deviation); M (min:max) = Median (min:max);

Figure 1: Sex and age groups distribution (| mean, ¦ median).

Figure 2: Residence distribution.

2.3.2 Alcohol

Only 27.2% of tha victims had consumed alcohol. Of them, the average concentration was 1.65 ±0.97, with values ranging from 0.3 to 4.

Table 2: Blood alcohol.

Variable

Details

Total

N=

81

Blood alcohol

22 (27.2%)

Blood alcohol level

μ ±SD

1.65 ±0.968

M (min:max)

1.35 (0.3:4)

μ ±SD = Mean (standard deviation); M (min:max) = Median (min:max);

There were no differences by gender, neither regarding alcohol consumption (25% of women and 27.9% of men, OR=0.86 [0.27, 2.74], p>0.999), nor the concentration in those who consumed alcohol (1.60 ±0.636 in women and 1.66 ±1.06 in men, p=0.907).

Table 3: Blood alcohol, by sex.

Variable

Details

F

M

Total

Statistics

Sex

20 (24.7%)

61 (75.3%)

81

Blood alcohol

5 (25.0%)

17 (27.9%)

22 (27.2%)

OR=0.86 [0.27, 2.74] (p>0.999)

Blood alcohol level

μ ±SD

1.60 ±0.636

1.66 ±1.06

1.65 ±0.968

T-test: p=0.907

M (min:max)

1.5 (1:2.4)

1.3 (0.3:4)

1.35 (0.3:4)

Blood alcohol level (incl. 0)

μ ±SD

0.40 ±0.768

0.463 ±0.929

0.447 ±0.888

MW: p=0.818

M (min:max)

0 (0:2.4)

0 (0:4)

0 (0:4)

μ ±SD = Mean (standard deviation); M (min:max) = Median (min:max); MW = Mann-Whitney Test; OR = odds-ratio [95% CI] and p value from Fisher test);

Age was not correlated to neither alcohol consumtion (T-test p=0.118), nor the concentration of blood alcohol (R=-0.297, p=0.180).

Table 4: Blood alcohol, by age.

Subset

N

Mean ±SD

Med (min:max)

Age (years) (Shapiro-Wilk normality test: p=0.007)

(total)

81 (100.0%)

58.36 ±18.8

61.0 (14.0:89.0)

Blood alcohol (Wilcoxon rank sum test with continuity correction: p=0.118)

yes

22 (27.2%)

53.95 ±17.1

55.0 (18.0:84.0)

no

59 (72.8%)

60.00 ±19.3

66.0 (14.0:89.0)

Blood alcohol level (Spearman's rank correlation rho: R=-0.297, p=0.180)

(total)

22 (27.2%)

1.65 ±1.0

1.4 (0.3:4.0)

Figure 3: Blood alcohol (| mean, ¦ median).

2.3.3 Cause of death

Most patients died due to head trauma (71.6%), followed by polytrauma (19.8%), asphyxia (5%) and head & neck trauma (3.7%). There seems to be no significant diference by sex (V=0.08, p=0.926).

Table 6: Cause of death, by Sex.

vs. Sex

Cause of death

F

M

(total)

head trauma

15 (25.9% / 75.0%)

43 (74.1% / 70.5%)

58 (71.6%)

polytrauma

3 (18.8% / 15.0%)

13 (81.2% / 21.3%)

16 (19.8%)

asphyxia

1 (25.0% / 5.0%)

3 (75.0% / 4.9%)

4 (4.9%)

head and neck trauma

1 (33.3% / 5.0%)

2 (66.7% / 3.3%)

3 (3.7%)

(total)

20 (24.7%)

61 (75.3%)

81 (100%)

V=0.08 (p=0.926)

Figure 4: Cause of death, by Sex.

2.3.4 Definite diagnosis

After examination the most comomon diagoses were intracranial hemorhage (64.2%) and cranial fractures (24.7%). There seems to be no significant diference by sex (V=0.19, p=0.730).

Table 8: Definite diagnosis, by Sex.

vs. Sex

Definite diagnosis

F

M

(total)

intracranial hemorhage

15 (28.8% / 75.0%)

37 (71.2% / 60.7%)

52 (64.2%)

cranial fractures

4 (20.0% / 20.0%)

16 (80.0% / 26.2%)

20 (24.7%)

atlanto-occipital fracture

1 (33.3% / 5.0%)

2 (66.7% / 3.3%)

3 (3.7%)

fractures

0

2 (100% / 3.3%)

2 (2.5%)

polytrauma

0

2 (100% / 3.3%)

2 (2.5%)

others

0

2 (100% / 3.3%)

2 (2.5%)

(total)

20 (24.7%)

61 (75.3%)

81 (100%)

V=0.19 (p=0.730)

Figure 5: Definite diagnosis, by Sex.

2.3.5 Circumstances

Most people died in road accidents (61.7% overall, 60.0% among women and 62.3% among men), in streets or public places (19.8% overall, 15.0% among women and 21.3% among men) or in hospitals (7.4% overall, 25.0% among women and 1.6% among men). All other circumstances were only found in men: train accident (3.7%), fall (4.9%) and home or work accident (1.2% each). There seems to be a significant relation to sex (V=0.42, p=0.026).

Table 10: Circumstances, by Sex.

vs. Sex

Circumstances

F

M

(total)

road accident

12 (24.0% / 60.0%)

38 (76.0% / 62.3%)

50 (61.7%)

street

3 (18.8% / 15.0%)

13 (81.2% / 21.3%)

16 (19.8%)

hospital

5 (83.3% / 25.0%)

1 (16.7% / 1.6%)

6 (7.4%)

train accident

0

3 (100% / 4.9%)

3 (3.7%)

fall

0

4 (100% / 6.6%)

4 (4.9%)

home

0

1 (100% / 1.6%)

1 (1.2%)

work accident

0

1 (100% / 1.6%)

1 (1.2%)

(total)

20 (24.7%)

61 (75.3%)

81 (100%)

V=0.42 (p=0.026)

Figure 6: Circumstances, by Sex.

2.3.6 Nature

Most patients died due to vehicle related causes (64.2%), followed by falls (27.2%) or violence (6.2%). There seems to be no significant diference by sex (V=0.13, p=0.708).

Table 12: Nature, by Sex.

vs. Sex

Nature

F

M

(total)

vehicle related

12 (23.1% / 60.0%)

40 (76.9% / 65.6%)

52 (64.2%)

fall

7 (31.8% / 35.0%)

15 (68.2% / 24.6%)

22 (27.2%)

violence

1 (20.0% / 5.0%)

4 (80.0% / 6.6%)

5 (6.2%)

other

0

2 (100% / 3.3%)

2 (2.5%)

(total)

20 (24.7%)

61 (75.3%)

81 (100%)

V=0.13 (p=0.708)

Figure 7: Nature, by Sex.

2.3.7 Other lesions

Most commonly associated lesions included cranial fractures (36.6%) and other fractures (32.4%); internal hemorrhage (35.2%), cerebral hemorrhage (8.5%) and other hemorrhages, mostly external (11.3%); polytrauma (22.5%). None of these seemed signigicantly associated to gender.

Table 14: Other lesions, by sex.

Variable

F

M

Total

Statistics

Sex

20 (24.7%)

61 (75.3%)

81

cranial fractures

5 (31.2%)

21 (38.2%)

26 (36.6%)

OR=0.74 [0.22, 2.42] (p=0.771)

other fractures

5 (31.2%)

18 (32.7%)

23 (32.4%)

OR=0.93 [0.28, 3.10] (p>0.999)

internal hemorrhage

3 (18.8%)

22 (40.0%)

25 (35.2%)

OR=0.35 [0.09, 1.36] (p=0.146)

cerebral hemorrhage

1 (6.2%)

5 (9.1%)

6 (8.5%)

OR=0.67 [0.07, 6.16] (p>0.999)

other hemorrhage

1 (6.2%)

7 (12.7%)

8 (11.3%)

OR=0.46 [0.05, 4.02] (p=0.673)

polytrauma

3 (18.8%)

13 (23.6%)

16 (22.5%)

OR=0.75 [0.18, 3.03] (p>0.999)

lacerations

3 (18.8%)

5 (9.1%)

8 (11.3%)

OR=2.31 [0.49, 10.94] (p=0.368)

contusion

1 (6.2%)

4 (7.3%)

5 (7.0%)

OR=0.85 [0.09, 8.19] (p>0.999)

others

0

1 (1.8%)

1 (1.4%)

OR=1.10 [0.04, 28.33] (p>0.999)

OR = odds-ratio [95% CI] and p value from Fisher test);

Figure 8: Other lesions.

2.4 Discussion

2.5 Conclusions

3 References

  1. R Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.