a. What is the null hypothesis?
b. Does the table provide any evidence of an association between the type of schedule and the rate of ‘serious medical errors’? Comment on the rate ratio, confidence interval, and P value.
c. What type of errors does the intervention schedule most effectively reduce? Justify your answer.
d. Why was it important that the researcher who decided whether a PRHO was allocated to either the intervention or traditional schedule had no knowledge about the PRHO? What do we call the design procedure employed in this type of study to prevent this potential bias? If this does not occur how could the results have been biased.
e. What is meant by ‘blinding’ and did it occur in this study? If a study is unblinded what might occur? How could this have affected the results in this study?
f. What other type of bias may effect this type of study design? Did it occur in this specific study?
g. In addition to the potential biases mentioned in questions d–f, what other explanations should be considered before concluding that the intervention schedule reduces the rate of serious errors? How likely are each of these?
h. Assuming the association between work schedule and serious errors is true, why is it important to look at the association between work schedule and preventable adverse events before any policy changes are made. What else should also be considered?
i. Does the intervention schedule lead to fewer preventable adverse events? Please comment on the rate ratio, confidence interval and P value? How precise is the estimate of the rate ratio?
j. How generalisable are the findings? Would you suggest implementation of the intervention schedule into the UK?
Q2 Blinding in RCTs
For each of the following trial designs, state whether it is possible to blind patients and/or researchers to the allocation of treatments
a. A drug trial of a new antidepressant compared to placebo where the outcome measure is a patient rated depression scale
b. Group versus individual speech therapy sessions for patients who have a language difficulty (dysphasia) after a stroke where the outcome measure is a recorded conversation reading a specific piece of text
c. A hyptertensive trial where patients are given a drug (beta-blocker) that also slows down heart rate or an identical looking placebo and a nurse measures the blood pressure.
d. A RCT of two different joint prostheses for osteoarthritis of the hip with the outcome measure being hip flexion without pain done by the researcher and a patient administered quality of life score.
e. A trial of acupuncture for arthritic pain using a patient completed pain index and participants who have never previously had acupuncture.
Q3 Economic evaluation
The economic evaluation was conducted as part of a randomised controlled trial which examined the effectiveness, acceptability and accessibility of a general practitioner with special interest (GPSI) dermatology service compared with routine hospital outpatient care. Patients who were referred to a hospital outpatient dermatology clinic and were deemed suitable to be managed by a general practitioner with special interest were randomised to either usual care (i.e. hospital outpatient care) or the GPSI service.
Resources were measured for the patients for nine months following randomisation. Most of the NHS resources were measured through computerised systems. Resources used by patients and their companions, and information on time off work (lost production) were measured through patient self completed postal questionnaires.
Two types of evaluation using differing perspectives were conducted.
