Skip to main content

Table 1 Dislocation outcome data reported for THA (n/N) vs HA (n/N) in 3 reviews

From: A case study of binary outcome data extraction across three systematic reviews of hip arthroplasty: errors and differences of selection

Study

BMJ

Cochrane

HTA

Comments

Dorr 1986 [22]

7/39 vs 2/50

7/39 vs 2/50

7/39 vs 2/50

Identical

Skinner 1989 [17]

NR

10/80 vs 11/100

11/89 vs 10*/91

Skinner and Ravikumar only report percentages rather than event data and only Ravikumar reports numbers in each trial arm (denominator)

(1-year data)

   
    

Errors:

Ravikumar 2000 [18]

18/91 vs 12/89

NR

18/89 vs 12/91

BMJ denominators for the 2 groups are the wrong way round; Cochrane generates its own denominators having failed to identify Ravikumar 2000 (follow-up to Skinner)

(13-year data)

   

Selection difference:

Numerators are all incorrect due to calculations based on percentages and incorrect denominators.

Baker 2006 [12]

3/40 vs 0/41

3/40 vs 0/41

3/40 vs 0/41

Identical

Keating 2006 [11]

3/69 vs 3/111

3/69 vs 2/69

3/69 vs 2/69

Selection difference:

BMJ alone analyses HA data from a separate trial arm (with 111 participants), but these data from this arm arguably should not be included in this analysis because different eligibility criteria were being applied (i.e. the surgeons and centres involved were either unwilling or unable to have participants randomised to THA).

Blomfeldt 2007 [9]

0/60 vs 0/60

0/60 vs 0/60

0/60vs 0/60

Identical

Macaulay 2008 [23]

1/17 vs 0/23

1/17 vs 0/23

1/17 vs 0/23

Identical

Mouzopoulos 2008 [8]

NR

NR

NR

NA

 

2/24 errors = 8 %

2/24 errors = 8 %

 

6 analysed studies = a/B vs c/D = 24 variables

4/24 selection differences =17 %

2/24 selection differences =8 %

  1. *Liang has 8/91 here: an error. Otherwise Liang has the same data as the HTA. THA, Total Hip Arthroplasty; HA, Hemiarthroplasty; NR, Not Reported; NA, Not Applicable. Findings are given in bold.