Reading a results table from a blood pressure study
NMC OSCE practice station

Evidence-based practice · 1 min to read · 10 min to perform

Open this station

A written station. Free to try, no sign-up.

The task

You are a practice nurse in a GP surgery. Your lead nurse has asked you to read the study summary in the folder, which tests home blood pressure monitoring with nurse telephone support, and to say what it means for your clinic. In the response sheet, explain what the results table shows and how far the results can be trusted. Then say what you would recommend for your surgery and what else you would want to find out.

Answer for this station

What a strong answer does

  • Reads the table accurately: both groups at 156 mmHg at baseline; at 6 months 142 compared with 148 mmHg, a difference of 6 mmHg lower with home monitoring.
  • Interprets the confidence interval and p value correctly: the interval -10.4 to -1.6 does not include zero and p = 0.008, so the difference is unlikely to be chance.
  • Distinguishes statistical significance from clinical importance: 6 mmHg is a modest average reduction, and the interval shows the true effect could be small.
  • Reports the target outcome correctly: 41 of 88 (47 %) compared with 29 of 91 (32 %) reached below 140/90 mmHg.
  • Notices that results are for completers only: 179 of 200 analysed and 21 withdrew, which could bias the findings.
  • Names strengths and other limits: random allocation, blinded outcome measurers; not blinded patients, short follow-up, no stroke or heart outcomes, only six surgeries.
  • Safety: Notes the adverse effect signal: dizziness in 5 compared with 2 patients, and says to advise patients.
  • Judges that the authors' conclusion about offering it to all adults goes beyond the data.
  • Gives a proportionate recommendation and next steps: offer to suitable patients, audit, check NICE hypertension guidance and look for longer-term trials.

Model answer

This is a randomised trial of 200 adults, with 100 in each group, comparing home blood pressure monitoring with nurse telephone support against usual care. The table shows that both groups started at an average systolic pressure of 156 mmHg. At six months the home monitoring group averaged 142 and the usual care group 148, so the home group was 6 mmHg lower. The 95 percent confidence interval is -10.4 to -1.6 and does not include zero, and p is 0.008, so the difference is unlikely to be due to chance. More people reached the target of below 140/90: 41 of 88, or 47 percent, compared with 29 of 91, or 32 percent. I need to say that statistical significance is not the same as importance. Six mmHg is a modest average reduction, and the true effect could be as small as 1.6 or as large as 10.4. Strengths are random allocation and that the nurses measuring blood pressure did not know the group. Limits are that only completers were analysed: 179 of 200, with 12 and 9 withdrawing, which may overstate the benefit. Patients could not be blinded, follow-up was only six months, the study did not look at strokes or heart attacks, and it was run in six surgeries with willing patients. Dizziness was reported by 5 compared with 2 patients. The authors' conclusion that all adults should be offered home monitoring goes further than the data. I would recommend offering home monitoring with nurse support to suitable patients, with advice about dizziness, not to everyone. I would check NICE hypertension guidance, audit our own results and look for longer-term trials.

What this station tests

  • Read each row of a results table with its group and time: baseline, follow-up and difference.
  • A confidence interval that does not include zero, with a small p value, suggests the difference is unlikely to be chance.
  • Statistical significance is not the same as clinical importance; look at the size of the effect and the interval.
  • Check who is missing from the analysis: withdrawals can make a treatment look better than it is.
  • Weigh benefits against harms and judge whether the authors' conclusion fits the data.

Sources for the clinical content

An original practice scenario with invented patients. Not an exam question. Written from the guidance above and reviewed before release: how we write and check stations.

Practise next

About the NMC OSCE →