Statistics for health care research
Give a statistical for health care research?
Questions to be graded
1) Read Exercise 31 in Statistics for health care research: a practical workbook.
2) Complete the study questions about the reading.
3) Check your answers to the study questions.
4) Copy and paste the Exercise 31: Questions to be Graded from page 236 into a word document.
a) Complete the questions on the document and place the document in the assignment drop box.
Or, as an alternative, you can
b) Print the page out, and complete the questions by hand (make sure your answers are legible). Scan your completed paper and place the document in the assignment drop box.
5) In order to receive full credit on calculated answers, please show your work. (Use Word’s equation editor, etc., and/or provide a short written description as to how you obtained the final result.)
6) Submit the assignment to the instructor by the end of Module 3.
The two groups whose results are reflected in tables 2 and are:
a) The prettest group and
b) The post test group.
The conditions of patients are compared to conditions after administering treatment. The group is subjected to health promotion programs. The differences before and after treatment is compared in order to assess the impact on subjects of study.
The t= 4.14 for total risk score in table 2 represents the greatest relative differences between the prettest and 3 months results. This t – value have got an asterisk (*) next to it in table 2, making it statistically significant. This asterisk in meant to direct the reader to the footnotes where the asterisk is said to represent p<0.05 which is the least acceptable value for statistical significance.
The t= 1.03 for health responsibility in table number 3 represents the smallest relative difference between pretest and 3 months outcomes. Since the t- value has no asterisk adjacent to it in table 3, it means that it is not statistically significant at 5% level of significance which is the least accepted value for statistical significance. This asterisk is supposed to guide the reader to the footnotes at the bottom of the table where asterisk is said to represent p<0.05.
Assumptions for conducting a t- test for dependant groups
a) The distribution of the scores is normally or approximately normally distributed
b) The dependent variables are measured at intervals or ration levels.
c) The groups being examined for differences are dependant based on matching or subjects serving as their own control.
d) The differences between the paired scores are independent
In this study, assumption number c was met
This is because the group of study was matched for their age, sex and and diagnosis. In terms of age, the whole 21 subjects were old women with an average mean age of 77 years.
In terms of matching diagnosis, 90% of the group was diagnosed with one or more chronic illness. This also applies to the matching of their health status.
In comparing the 3 and 6 months outcome for exercise variable in table we note that
a) The t-value reduced significantly from t=-7.75* to t=-3.93*
b) That in both cases the results were statistically significant
c) There was reduction in mean standard deviation from 2.58 to 2.29
The large t ratio of t=-7.75* indicates large differences in exercise level from pretest to 3 months which was statistically significant.
Reduced to t=-3.93 implying that though there was a statistically significant reduction in differences in exercise level among the sample group in the in the long run
These results also show that although there was initial increase in mean exercise by the sample group in the first three months (from 1.88 to 2.58), there was a significant reduction in their mean exercise (from 2.58 to 2.29) in the period covering 3 to 6 month period meaning that health promotion programs did not have a positive long term effect on exercise subject since it showed a reduction in exercise participation verses the required increases in exercise participation.
Value t= 2.03* for cholesterol variable is the smallest listed significant t- ratio in table 2.This is because it has asterisk (*)next to it which is used to indicate the t values which are significant as it also directs the reader to the p values less than 0.005 at the bottom of the table.
This also means that cholesterol is the variable which has less statistical significance on the sample group as compared to other variables that are statistically significant.
Larger t ratios are likely to be statistically significant because of the smaller p values that are likely to be observed hence the higher likelihood of rejecting the null hypothesis.
The smaller p values reduces the significance level for the test statistic hence the likelihood of accepting the null hypothesis.
Larger t ratios indicates a higher differences between the control and experimental group in terms of variables being tested which means that the treatment being administered has a higher effects on the study group hence higher level of significance and high probability value and high chances of rejecting the null hypothesis that there is no relationship between cholesterol level and health intervention programs.
The findings of this study indicate that health intervention programs did not have a statistically significant effect on Systolic BP. This is shown by the absence of asterisk next to its t – value in table number 2 either 3 months or 6 months column. it has t=1.57 and t=1.66 at 3 and 6 months respectively with no asterisk next them.
Mean of systolic BP
At pretest = 121.7
At 3 months = 117.2
At 6 months = 115.6
Standard deviation of systolic BP
At pretest = 14.6
At 3 months = 12.3
At 6 months = 13.4
There is a significant decline in the mean value for the systolic BP from pretest to 3 months to 6 months. This is clinically important because it indicates a reduction in blood pressure for the sample patients.
A high blood pressure of more than 140 is a health risk and leads to a condition called isolated systolic hypertension, (Wills, et al, 2011, p.12). The study have also found that hypotension in most cases have no obvious symptoms but it can lead many of life threatening cardiovascular diseases such as heart attach and stroke.
Normal adult blood pressure should be below 130milimetres of mercury hence the need to reduce this to within those levels to minimize associated cardiovascular risks.
The design for this research is not strong in the sense that it used one single group which made it difficult to determine the effects of treatment without comparing with a separate control group. Comparisons between the experimental and controlled group is necessary for carrying out a research in order to make a significant conclusion
As a health care provider I will implement this health intervention programs at my facility. However, I will re design my study to include the controlled group which will not be subjected to this treatment. This will assist me in drawing a conclusive comparison between experimental and controlled group at any given time of my study.
D) Is the sample mean sufficient to conclude that the treatment has a significant effect? Give reasons for your decision.
A sample mean of 25 is sufficient since is recommended for using student t – test hence we can confidently conclude that the treatment has a significant effect.
E) If the size of the sample was changed to n = 16, would the sample mean be sufficient to conclude that the treatment has a significant effect? Give reasons for your decision.
Reducing n to 16 will increase the SE (M) from 3 to approximately 3.7. Reduced standard error mean will in turn increasing the z value for the study. This will reduce the rejection zone hence increasing the confidence level making the statistical result more significant. Therefore reducing n value to 16 means the sample is sufficient to conclude that the treatment has significant effect, (MacCallum, 1996).
This research was conducted to determine the effects of health promotion programs on cardiovascular risk factors, health behaviors and life satisfaction in institutionalized old women, (Kim, et al 2003).
The findings of this study revealed that the health promotion programs have positive effects on reducing total risk score, improvement health behaviors and life satisfaction for the sample group.
The use of single group of elderly women in carrying out this study is the major limitations because for study to be successful there is need to compare the experimental group with the controlled group.
I recommend the use of this method in many health care provider institutions but with a redesigned sample consisting both the experimental and controlled sample group.
Wills AK, Lawlor DA, Matthews FE, Aihie Sayer A, Bakra E, et al. (2011) Life Course Trajectories of Systolic Blood Pressure Using Longitudinal Data from
Eight UK Cohorts. PLoS Med 8(6): e1000440. doi:10.1371/journal.pmed.1000440
MacCallum, Robert C, Browne, Michael W, Sugawara, Hazuki M.(1996):
Power analysis and determination of sample size for covariance structure modeling.
Psychological Methods, Vol 1(2), Jun 1996, 130-149. doi: 10.1037/1082-989X.1.2.130
Kim C. J and Song R. (2003): Effects of a health promotion on cardiovascular risk factors, health factors, and life satisfaction in institutionalized elderly women. International Journal of nursing studies, 40(4), 375-81.