Correlates of Length of Hospital Stay among Emergency General Surgery Patients

Correlates of Length of Hospital Stay 1
Correlates of Length of Hospital Stay among Emergency General Surgery Patients
Name
Course Number
Date
Faculty Name

Correlates of Length of Hospital Stay 2
Correlates of Length of Hospital Stay: Retrospective Analysis of Emergency General Surgery
Admission between 2011 and 2017
Abstract
Emergency general surgery is very crucial in saving life. Further, bed space in hospitals is a
scarce resource, especially in low and middle income countries. Therefore, having a clear
understating of the causes of longer hospital stay might be helpful in mitigating this issue hence
improving approaches of dealing with emergency surgery admissions. To achieve this,
retrospective data on emergency general surgery admissions obtained between 2011 and 2017
has been used for the analysis. Using non-parametric bivariate analysis methods such as
Wilcoxon and Kruskal-wallis test, we determined the significant correlates of length of hospital
stay among emergency general surgery patients. We found that age, being admitted as a result of
transfer, experiencing complications during surgery, a case being operative and surgery
involving endoscopic procedures were statistically significant correlates of hospital stay.
However, there was no significant differences in length of hospital stay between patients who
died and who survived.
Introduction
Length of hospital stay is a marker of severity of a medical condition. Patients who stay longer in
the hospital would more likely to have been admitted with a more severe condition than their
counterpart or have suffered hospital acquired infections during admission. Although these might
be the only correlates of length of hospital stay, there could be other significant factors. Hospital
admissions target patients whose conditions are better when managed by professional healthcare
providers other than homecare. The patients are discharged after a specialised has confirmed that
the state of the patient can be managed in a home environment. In most occasion, drug
prescriptions are accompany hospital discharge for the patient to use until full recovery.
Occasionally, readmission are also observed, mostly due to mismanagement, co-infections or
even new infections.
Emergency general surgery is performed in cases where a patients is in a bad state and the
procedure may not be managed for a long time. This type of procedure might be as a result of an
accident or a medical condition that has just been realised. In some cases, it might be due to
inevitable pain, which can only be managed by a surgical procedure. Further, it might be very
hard to evaluate how good a surgical procedure was conducted. Therefore, using the post-
operative characteristics of a patient such as the length of hospital can be very instrumental in
trying to understand how good a procedure was conducted and vice versa. Understanding the
factors contributing to a longer stay in the hospital after surgery, more could be done on the
potential confounders of wellness, hence improving the general state of health for the patients
undergoing general surgical procedures.
Despite the intensive hospital care and the good intentions of the emergency surgery procedures,
some patients end up dying in the hospital, either during the procedure of after or after discharge.
However, post-operative deaths constitutes of a small percentage of the overall deaths endured
within the surgical service (Heeney et al., 2014). In some other cases, patients are transferred to
other hospital due to a couple of reasons such as severity of the condition – either to a less or
more comprehensive hospital. This paper focus on the analysis of retrospective general surgery
admissions data to determine the factors associated with a longer hospital stay.

Correlates of Length of Hospital Stay 3
Data
Data used in the paper is obtained from a retrospective study for all emergency general surgery
admissions from February 1, 2011 to December 31, 2017. It comprises of patient demographic
data, comorbidities, hospital length of stay, admission diagnoses, surgical interventions, in-
hospital complications and mortality information. The data has been categorised by age, for
younger and older than 60 years of age. For this study, length of hospital stay is used as the
outcome variable and it is measured in days (the arithmetic difference between date of hospital
discharge and date of admission). Some patients had 0 days of hospital stay, indicating that their
surgical procedures were minor and there was no medical need of being admitted. The data
constitutes of 15,297 observations and 25 variables.
Methods
In the analysis, R Studio version 1.2.1335 was used to implement the test statistic data science
methods used (RStudio Team, 2018). Non-parametric tests Wilcoxon and Kruskal-wallis test
were used to test significance in between the length of hospital stay and the dependent variables .
Results and Discussions
On average, males had a longer length of hospital stay compared to the female patients based on
the mean values. There was no difference in median days of hospital stay after surgery. On a
case analysis, there was a male patient who stayed in the hospital for over a year. The diagnosis
information recorded for this patient was constipation, he was admitted as an emergency, did not
incur any complications and he lived. The patient distribution was approximately balanced by
gender with 55% of them being females. The overall average age of the patients at admission
was 47.3 years, while male patients were slightly older. Approximately 70% of the patients were
above 60 years of age and a similar distribution observed for both genders. There was a very
high survival rate and approximately 0.4% mortality was reported.
Table 1: A table of descriptive statistics stratified by gender
Female
(n=8403)
Male
(n=6894)
Overall
(n=15297)
Length of Hospital Stay
Mean (SD) 2.92 (4.20) 3.14 (7.12) 3.02 (5.70)
Median [Min, Max] 2.00 [0.00, 94.0] 2.00 [0.00, 456] 2.00 [0.00, 456]
Gender
Female 8403 (100%) 0 (0%) 8403 (54.9%)
Male 0 (0%) 6894 (100%) 6894 (45.1%)
Age of the Patient at Admission
Mean (SD) 46.1 (21.5) 48.7 (21.3) 47.3 (21.5)
Median [Min, Max] 41.0 [16.0, 103] 45.0 [15.0, 103] 43.0 [15.0, 103]
Age above or below 60 years