Describe the results of this analysis for frequent and moderate ‘treatment and release
use’ of the ED. Report on the strength and direction of the statistically significant
associations.

Chi square test and t-test

Order Description

ASSESSMENT 1 – 40%
Assessment 1 must be completed in 900-1200 words and is worth 40 marks. While references
are not required, where you use other authors’ work, this should be appropriately referenced.
References and in-text citations must not be included in the word count. You are required to
use the Assessment 1 Template (available on the Unit website).
You are required to answer all components of each of the four Parts detailed below. The
Assessment 1 paper should predominantly be written in narrative form. Mark allocations are
shown throughout. Mark allocations provide a guide to the word length for each Part (on
average, 10 marks is equivalent to a maximum of 300 words). Marks include a 20%
component for expression (including spelling and grammar), referencing (if used) and
adherence to the minimum and maximum word count.
Part 1 (10 marks)
The following questions relate to a randomised controlled trial to examine the effects of
applying a tailored delirium preventive protocol to ICU patients. Participants were
randomised to an intervention group or to a control group receiving usual care.
Table 1 shows a comparison of baseline characteristics of participants. Chi-square and t-tests
were used for comparing the demographics and clinical characteristics between the
intervention and the control groups.
9058: Evidence Based Health Care
Semester 2, 2015 2
Assessment items
1.1 In Table 1, the Chi-square test and t-test were used to determine the significance of
differences between categorical and numerical variables of two independent groups (the
intervention and the control group). Which variables would the Chi-square test have
been applied to, and which would the t-test have been applied to? [3 marks]
1.2 Which of these tests is the parametric statistical test and what are the main assumptions
about the data being subjected to that test? [4 marks]
1.3 Describe the results of the baseline comparisons in Table 1. [3 marks]
Part 2 (10 marks)
The following table shows some of results of multinomial logistic regession analyses of
predictor variables associated with 5650 healthy adults’ frequent or moderate ‘treatment and
release’ use of an Emergency Department. The predictor variables are bivariate (Yes/No or
Female/Male) except for the numerical variables ‘Age’ and ‘Number of outpatient visits’.
You are asked to focus just on the odds ratios and confidence intervals (OR and 95% CI).
Assessment items
2.1 Describe the results of this analysis for frequent and moderate ‘treatment and release
use’ of the ED. Report on the strength and direction of the statistically significant
associations. [5 marks]
2.2 Consider the situation where other studies have produced similar results for the main
factors predicting frequent and moderate ‘treatment and release use’ of the ED (NB –
you don’t need to try and find any other studies). What might be the implications of this
evidence for clinical practice when providing nursing services for such patients? [5
marks]
9058: Evidence Based Health Care
Semester 2, 2015 3
Part 3 (10 marks)
The following table shows results of a study that linked hospital facility data with staff nurse
survey data (N = 1024) and surgical patient discharge data (N = 76,036) from 14 teaching
hospitals with 700 or more beds. Logistic regression models that adjusted for a range of
patient and hospital factors were used to estimate the effects of three nursing characteristics
on risk-adjusted patient mortality within 30 days of admission.
The three nursing characteristics were nurse staffing (patients per nurse), nurse work
environment (better vs. mixed or poor) and nurse education (the percentage of nurses with a
Bachelor degree in nursing).
The fully adjusted logistic regression model controlled for patient age, gender, transfer status,
type of admission, major diagnostic category and co-morbidities, as well as hospital location
and number of beds.
Assessment items
3.1 Describe the results of this analysis, focussing on the statistically significant predictors
in the fully adjusted model. Report on the strength and direction of the associations. For
the purposes of this discussion, consider the fully adjusted result for nurse staffing
(patients per nurse) to be statistically significant. [5 marks]
3.2 Consider the situation where other studies have produced similar results for the main
nursing characteristics predicting patient mortality (NB – you don’t need to try and find
any other studies). What might be the implications of the evidence for the health system
in relation to the quality and safety of patient care? [5 marks]