Md2 Block4 Research Part 2 Chapt 37, 38, 40

Front 1

37
Life style variations in three INTERSALT centres (Beijing, Nanning and Tianjin) in the People's Republic of China compared with 45 other INTERSALT centres, standardised for age/sex/centre.
Variable 3 PRC centres 45 other centres
24h sodium excretion (mmol, mean) 206.3 (70.9) 163.0 (62.8)
Alcohol drinkers (%) 31.8 61.1
Alcohol in drinkers (ml/week, mean) 68.6 (142.7) 197.8 (269.9)
Body mass index (kg/m2, mean) 22.7 (2.9) 25.4 (4.1)

Back 1

The title of the table is satisfactory.

The columns of the table are adequately labelled.

The entries in the table for 24 hour sodium excretion are satisfactory and sufficient.

Front 2

38
A simple routine test for the presence of HIV was carried out on 300 high risk subjects (intravenous drug users). A more accurate but time-consuming and expensive test (the 'gold standard') was also carried out on the subjects to assess the accuracy of the routine test. The following results were obtained:
HIV by 'gold standard'
Yes No Total
Routine test +ve 92 10 102
-ve 2 196 198
Total 94 206 300

Back 2

The estimated prevalence of HIV in the relevant population is 0.31

The negative predictive value of the test is 99.0%.

Front 3

Properties of diagnostic tests

Back 3

For a condition which is easily treatable, we should like the diagnostic test to have a high sensitivity.

The likelihood ratio for a positive diagnostic test result is the ratio of the chance of a positive result if the patient has the disease to the chance of a positive result if he/she does not have the disease.

Front 4

Use of screening tools for Chlamydia

The potential of the ligase chain reaction as a screening tool for the detection of Chlamydia was assessed in 890 women aged 18-35 years (Grun L et al. BMJ 1997; 315: 226-230). The prevalence of confirmed Chlamydia infection was 2.6% (95% CI 1.6 to 3.6%) in all women. The prevalence of Chlamydia on the basis of enzyme immunoassay was 1.6% (0.8% to 2.7%), with a sensitivity of 60% and a specificity of 100%. The prevalence on the basis of ligase chain reaction was 2.5% (1.5% to 3.9%) with 90% sensitivity and 99.8% specificity. Select all of the following statements which you believe to be true.

Back 4

A greater number of false positives are identified with the ligase chain reaction.

No false positives were detected using the immunoassay method.

Front 5

PCR methods for detecting viral infection of CNS

In a study to assess the usefulness of PCR methods to detect viral infections of the central nervous system (CNS), CNS samples from 410 patients were sent to a diagnostic laboratory to be examined by nested PCR for viruses associated with CNS infections (Jeffery KJ et al. Lancet 1997; 349: 313-317). Clinical classification of viral infection of the CNS (assumed to be the gold standard) was available for all 410 patients, of whom 12 (2.9%) were classified as cases of definite viral infection of the CNS, 51 (12.4%) were classified as probable cases, seven (1.7%) as possible cases and 340 (83.0%) were judged not to have a viral infection of the CNS. A virus was detected by PCR in 10 of the 12 definite cases, 10 of the 51 probable cases, none of the possible cases and two of the no infection cases. A diagnosis of definite viral infection of the CNS was 88.2 times as likely to occur in a patient with a positive PCR result as in a patient wit

Back 5

If all probable and possible viral infections were classified as viral infections, the specificity of a positive PCR result would be 99.4%.

Front 6

39
Cohen's kappa

Back 6

We use Cohen's kappa to measure the agreement between two observers who assess disease severity using a categorical scale of measurement.

Front 7

40
Number needed to treat

Background Infections caused by multiply resistant Shigella species are a major cause of childhood morbidity and mortality in Third World countries. A randomized double-blind study was undertaken to test the effects of ciprofloxacin treatment in children with shigella dysentry.

Methods The efficacy and toxic effects of ciprofloxacin suspension (10mg/kg) every 12 h for 5 days, maximum individual dose 500mg) were compared with those of pivmecillinam tablets (15-20 mg/kg every 8h for 5 days, maximum individual dose 300mg). 143 children aged 2-15 years with dysentry of 72h or less duration were enrolled in the study. Patients stayed in hospital for 6 days, and were followed up 7, 30 and 180 days after hospital discharge. Joint symptoms and function were assessed daily for 6 days.

Back 7

Clinical success was defined as the absence of frank dysentry on day 3, and on day 5 no bloody-mucoid stools, one or no watery stool, six or fewer total stools, and no fever. If no shigella were isolated from faecal samples on day 3 or thereafter, treatment was judged bacteriologically successful. Secondary endpoints of interest were duration of fever, duration of abdominal pain, and the total number of stools for each patient in the study period.

Statistical analysis Differences in proportions were tested with the Chi-squared test with continuity correction or Fisher's Exact test if the predicted size of any cell was five or less. Differences in continuous variables were assessed with Student's t-test or the Mann-Whitney U test otherwise. All tests of significance were two-tailed.

Findings 13 patients were excluded since they did not meet eligibility criteria; 10 withdrew before day 5. Thus 120 patients (60 in each group) completed the study. Treatment was clinically successfu

Hint 7

in 48 of 60 patients who received ciprofloxacin and in 39 of 60 patients who received pivmecillinam (difference in percentages = 15.0%, 95% CI for difference = -0.7% to 30.8%, P=0.10). Treatment was bacteriologically successful in all of the patients who received ciprofloxacin, and in 54 of the patients who received pivmecillinam (difference in percentages = 10.0%, 95% for difference = 2.4% to 17.6%, P=0.03). Joint pain began in 18% of patients who received ciprofloxacin and 22% of patients who received pivmecillinam (P>0.2), and no patient had signs of arthritis. The median number of stools during the study was 35 in the ciprofloxacin group and 36 in the pivmecillinam group (P>0.2).

Interpretation Ciprofloxacin suspension and pivmecillinam had the same clinical efficacy in this trial. Ciprofloxacin had greater bacteriological efficacy and was not associated with the development of arthropathy. It can be concluded that ciprofloxacin is an effective and safe drug for use in multiply r

Front 8

Interpretation Ciprofloxacin suspension and pivmecillinam had the same clinical efficacy in this trial. Ciprofloxacin had greater bacteriological efficacy and was not associated with the development of arthropathy. It can be concluded that ciprofloxacin is an effective and safe drug for use in multiply resistant childhood shigellosis.

Back 8

The main effects of interest were the differences between the two treatments in the percentages of children who were judged clinically successful (15%) and biologically successful (10%).

The only result quoted that achieved statistical significance at the 5% level was the comparison of the percentages of children who were bacteriologically successful.

Front 9

2x2 test

Back 9

(a+c) / n

n = size of population

Slide 20 for formula

Front 10

Sensitivity
w/ disease Identified with test

Back 10

= a / (a+c)

proportion of individuals with the disease who are correctly identified by the test

Front 11

Specificity

Back 11

w/out disease Identified with test
d/(b+d)

proportion of individuals without the disease who are correctly identified by the test

Front 12

positive predictive value

Back 12

a/(a+b)

Front 13

Negative predictive value =

Back 13

d / (c + d)

= proportion of individuals with a negative test result who do not have the disease, and is calculated as

Front 14

diagnostic test
is used as a =
2ct

Back 14

As a staging

As a monitoring

Front 15

Positive Likelihood ratio.... same as Specificity eq

Back 15

w/out disease Identified with test
d/(b+d)

Front 16

negative Likelihood ratio ratio

A likelihood ratio can also be generated for a negative test result and is most easily calculated as

Back 16

(1 – sensitivity) /specificity.

Front 17

Number-needed-to-treat

In a randomized, placebo-controlled study of riboflavin for the prevention of migraine (Schoenen J et al. Neurology 1998; 50: 466-470), the authors reported that the proportion of patients who had 'responded' (improved by at least 50%) was 15% for placebo and 59% for riboflavin (P=0.002). The number-needed-to-treat for effectiveness was 2.3. Select all of the following statements which you believe to be true.

Back 17

The number-needed-to-treat for effectiveness is calculated as 1/(0.85-0.41).

The number-needed-to-treat for effectiveness is calculated as the reciprocal of the difference in the proportions of the two positive outcomes.

At least 3 patients need to be treated to ensure that one extra person will be a responder.

Front 18

Trial of treatment for septic shock

The summary of a double-blind randomized controlled trial of monoclonal antibody to human tumour necrosis factor in treatment of septic shock

Background: Despite the availability of potent antibiotics and intensive care, mortality rates from septic shock are 40-70%. We assessed the safety and efficacy of murine monoclonal antibody to human tumour necrosis factor alpha (TNF alpha MAb) in the treatment of septic shock.

Methods: In a randomized, multicentre, double-blind placebo-controlled clinical trial in 105 hospitals in the USA and Canada, we randomly assigned 1879 patients a single infusion of 7.5 mg/kg TNF alpha MAb (n=949) or placebo (0.25% human serum albumin n=930). Our main outcome measurement was the rate of all-cause mortality at 28 days.

Back 18

Findings: 382 (40.3%) of 948 patients who received TNF alpha MAb and 398 (42.8%) of 930 who received placebo had died at 28 days (95% CI -0.02 to 0.07, P=0.27). We found no association between therapy with TNF alpha MAb and increased rapidity in reversal of initial shock or prevention of subsequent shock. Similarly, baseline plasma interleukin-6 concentrations of more than 1000 pg/mL or detectable circulating TNF concentrations were not associated with improvement in survival after TNF alpha MAb therapy. Coagulopathy, but not other organ or system failures, was significantly decreased in the TNF alpha MAb group compared with placebo (day 7, P<0.001; day 28, P=0.005). Serious adverse events were reported in 55.2% of patients given placebo and 54.1% in the TNF alpha MAb group.

Interpretation: We did not find an improvement in survival after septic shock with TNF alpha MAb. Therapy not solely dependent on TNF alpha blockade may be required to improve survival.

Hint 18

The main effect of interest was precise.
Right answer.
This statement is true. Because of the large sample size, the standard error of the difference in the percentages dying is quite small and so the confidence interval for the difference in death rates between the two groups is quite narrow. The 95% CI is -2% to 7%, i.e. the improvement in survival rates could have been as low as 2% in favour of the placebo group or as much as 7% in favour of the treated group.

Front 19

Trial of treatment for septic shock

The summary of a double-blind randomized controlled trial of monoclonal antibody to human tumour necrosis factor in treatment of septic shock (Abraham E et al. Lancet 1998; 351: 929-933) is shown below.

Background: Despite the availability of potent antibiotics and intensive care, mortality rates from septic shock are 40-70%. We assessed the safety and efficacy of murine monoclonal antibody to human tumour necrosis factor alpha (TNF alpha MAb) in the treatment of septic shock.

Methods: In a randomized, multicentre, double-blind placebo-controlled clinical trial in 105 hospitals in the USA and Canada, we randomly assigned 1879 patients a single infusion of 7.5 mg/kg TNF alpha MAb (n=949) or placebo (0.25% human serum albumin n=930). Our main outcome measurement was the rate of all-cause mortality at 28 days.

Back 19

Findings: 382 (40.3%) of 948 patients who received TNF alpha MAb and 398 (42.8%) of 930 who received placebo had died at 28 days (95% CI -0.02 to 0.07, P=0.27). We found no association between therapy with TNF alpha MAb and increased rapidity in reversal of initial shock or prevention of subsequent shock. Similarly, baseline plasma interleukin-6 concentrations of more than 1000 pg/mL or detectable circulating TNF concentrations were not associated with improvement in survival after TNF alpha MAb therapy. Coagulopathy, but not other organ or system failures, was significantly decreased in the TNF alpha MAb group compared with placebo (day 7, P<0.001; day 28, P=0.005). Serious adverse events were reported in 55.2% of patients given placebo and 54.1% in the TNF alpha MAb group.

Interpretation: We did not find an improvement in survival after septic shock with TNF alpha MAb. Therapy not solely dependent on TNF alpha blockade may be required to improve survival.

Hint 19

At least 40 patients need to be treated in order to prevent one death from septic shock by 28 days.