Essay About Continuous Data And P-Value

Biostatistics Case
Essay Preview: Biostatistics Case
Report this essay
Age: Continuous
Urgency of operation
ordinal
Length of hospital stay
continuous
Type of surgery
nominal
Ejection fraction estimate
ordinal
Preoperative dialysis
nominal
Patients ages in each mortality groups
Group Statistics
Mortality Status
Std. Deviation
Std. Error Mean
65.1778
13.07279
.41909
72.4766
6.58434
1.29130
Independent Samples Test result
Levenes Test for Equality of Variances
t-test for Equality of Means
Sig. (2-tailed)
Mean Difference
Std. Error Difference
95% Confidence Interval of the Difference
Lower
Upper
Equal variances assumed
10.804
-2.836
-7.29885
2.57339
-12.34873
-2.24896
Equal variances not assumed
-5.376
30.535
-7.29885
1.35760
-10.06940
-4.52829
As age is a continuous data and mortality status is nominal data, it can be solved by using independent samples t-test
Hypotheses:
Null hypothesis: there is no difference of the mean ages between the 30-day mortality statuses.
Alternative hypothesis: two groups have different mean ages
Assumptions:
Ages in each group follows normal distribution.
Groups are independent.
Patients within each group are independent.
As equality of variances is 13.072792/ 6.584342 =3.9 > 2, the equal variances is not assumed.
T-score = difference in sample means/ se= -5.376
P- value:
By using SPSS data package, the p-value for t=-5.376 is less than 0.000 < 0.005, since the p-value is very small smaller than 0.005, we reject the null hypothesis and the difference in the mean of ages for alive and death is significant. Conclusion: the mean age for cardiac disease death is 72 which is higher than those of the alive group (65), as within the 95% CI (-10.069, -4.528), it does not include "0", and the p-value is much smaller than 0.005, thus the result suggests that the difference in ages can lead to significantly different mortality rate. The study results support that patients with higher age are more like to contribute to the 30-day mortality rate. Using SPSS data package, Length of hospital stay compared to the health status As they are asymmetrical data, we must use Kruskal-wallis test on SPSS, LOS as compared by different cardiac illness status Status Mean Rank Length of hospital stay Elective 392.02 Urgent 676.63 Emergency 634.45 Salvage 809.13 Total Descriptives Status Statistic Std. Error Length of hospital stay Elective 9.8117 .64413 95% Confidence Interval for Mean Lower Bound 8.5467 Upper Bound 11.0767 5% Trimmed Mean 8.3481 Median 7.0000 Variance 255.584 Std. Deviation 15.98699 Minimum Maximum 374.00 Range 374.00 Interquartile Range Skewness 19.503 Kurtosis 439.726 Urgent 17.5015 .87272 95% Confidence Interval for Mean Lower Bound 15.7847 Upper Bound 19.2183 5% Trimmed Mean 15.3013 Median 13.0000 Variance 252.105 Std. Deviation 15.87782 Minimum Maximum 154.00 Range 153.00 Interquartile Range 10.00 Skewness 4.594 Kurtosis 29.535 Emergency 24.1277 4.57375 95% Confidence Interval for Mean Lower Bound 14.9212 Upper Bound 33.3341 5% Trimmed Mean 19.1761 Median 14.0000 Variance 983.201 Std. Deviation 31.35603 Minimum Maximum 160.00 Range 160.00 Interquartile Range 18.00 Skewness 2.933 Kurtosis 9.399 Salvage 44.0000 29.05455 95% Confidence Interval for Mean Lower Bound -48.4645 Upper Bound 136.4645 5% Trimmed Mean 41.0556 Median 17.5000