Effectiveness of Antipsychotic Drugs in Patients with Chronic Schizophrenia
2005 · paper · Clinical trials
A double-blind trial randomly assigned 1493 people with schizophrenia to one older and four newer antipsychotic drugs for up to 18 months. Most stopped their assigned drug; olanzapine had the lowest discontinuation rate but more weight gain and metabolic changes.
2 findings
Substance Table 1, row 'Any combination including olanzapine, quetiapine, or risperidone': Total column prints "95 (7)" but the group counts 31, 32, 33, 21 and 8 sum to 125 (about 9%).
- p. 5 Table 1, Baseline antipsychotic medications, row 'Any combination including olanzapine, quetiapine, or risperidone', Total column
95 (7)
- p. 5 Table 1, same row, arm columns
31 (9)
- p. 5 Table 1, Baseline antipsychotic medications, 'Any combination...' row, Total column
31 (9) 32 (10) 33 (10) 21 (8) 8 (4) 95 (7)
Judge's note: Read Table 1 on the page 5 image. The arm counts sum to 125, not 95. The printed 95 (7) is identical to the 'Quetiapine alone' total two rows above. With 125 the six medication categories in the Total column sum to 1460 (the total N); with 95 they sum to 1430. Every other row's total equals the sum of its arms, and no footnote explains a different counting rule.
Second check: another model agreed
Report a problem with this finding
Substance Table 2, duration of successful treatment, quetiapine vs. risperidone: "1.30 (1.04–4.63)"; the upper bound is far out of line with the estimate, the lower bound and every neighbouring interval (e.g. "1.28 (1.00–1.64)").
- p. 9 Table 2, Duration of successful treatment, Quetiapine row, Risperidone column
1.30 (1.04–4.63)
- p. 6 Results, text
ratio, 0.77; P=0.021)
Only the authors can say what is right here.
Judge's note: Read the upright page 9 image; the cell prints 1.30 (1.04–4.63). The estimate is 1.25 times the lower bound, but the upper bound is 3.56 times the estimate. All other hazard-ratio intervals in the table are roughly symmetric on the ratio scale, including the adjacent 1.28 (1.00–1.64) and 1.06 (0.85–1.33). An upper bound near 1.63 would fit the same pattern, which suggests a first-digit misprint; the authors should confirm the value. The text on page 6 gives the reciprocal ratio 0.77 (1/1.30).
Second check: another model agreed
Report a problem with this finding
7 minor (typos and the like; most checked by the judge only)
Wording Successful treatment is defined as "a CGI Scale score of at least 3 (mildly ill) or a score of 4 (moderately ill) with an improvement of at least two points from baseline"; with higher scores meaning greater severity, "at least 3" reads the wrong way round.
- p. 4 Methods, Statistical analysis
a CGI Scale score of at least 3 (mildly ill) or a score of 4 (moderately ill) with an improvement of at least two points
- p. 9 Table 2, footnote ¶¶
a CGI severity score of at least 3 (mildly ill)
- p. 4 Methods, Statistical Analysis
patients had a CGI Scale score of at least 3 (mildly ill) or a score of 4 (moderately ill) with an improvement of at least two points
Only the authors can say what is right here.
Judge's note: Read page 4 and the Table 2 footnote on page 9; both use 'at least 3'. Page 3 states that CGI scores run from 1 to 7 with higher scores indicating greater severity. Taken literally, 'at least 3' covers scores 3 to 7 and makes the second clause redundant; '3 or less' appears intended. This is a wording ambiguity repeated in two places, and the authors should confirm the intended criterion.
Second check: another model agreed
Report a problem with this finding
Wording Text says anticholinergic drugs were added in quetiapine patients "3 percent vs. 8 to 10 percent", but Table 3 prints olanzapine as "25 (7)", so the table's range for the other groups is 7 to 10 percent.
- p. 7 Results, Concomitant medications
3 per-cent vs. 8 to 10 percent
- p. 13 Table 3 (continued), Medications added, Anticholinergic agents, Olanzapine
25 (7)
- p. 7 Concomitant medications
cent vs. 8 to 10 percent)
- p. 7 Concomitant medications
cent vs. 8 to 10 percent).
Only the authors can say what is right here.
Judge's note: Read page 7 text and Table 3 (continued) on page 13. The table gives olanzapine 7, risperidone 9, perphenazine 10 and ziprasidone 8. With the footnote denominator of 333, 25/333 is 7.51%, which rounds to 8 and would match the text; with N=336 it is 7.44%, which matches the table. The text and the table cell therefore disagree by one percentage point, and the authors should say which denominator applies. The same pattern appears in C011.
Second check: another model agreed
Report a problem with this finding
Wording Text: "gained 7 percent or more of their baseline body weight (30 percent vs. 7 to 16 percent)"; Table 3 row label: "Weight gain >7%".
- p. 7 Results, Weight Gain and Metabolic Changes
gained 7 percent or more of their
- p. 12 Table 3, Weight change from baseline to last observation
Weight gain >7% — no./total no. (%)
Judge's note: Read page 7 text and Table 3 on page 12. The threshold is described as 7 percent or more in the text and as strictly greater than 7% in the table, for the same percentages. This is a minor labelling inconsistency; the numbers agree (30 vs. 16, 14, 12, 7).
Second check: another model agreed
Report a problem with this finding
Wording Table 3 (continued), Cholestatin drugs, olanzapine: "15 (4)"; with the footnote denominator of 333, 15/333 is 4.5%, which rounds to 5.
- p. 13 Table 3 (continued), Cholestatin drugs, olanzapine
15 (4)
Only the authors can say what is right here.
Judge's note: Read page 13. The footnote gives 333 as the olanzapine denominator for medications added. 15/333 is 4.505%, which is not an exact tie and rounds to 5; 15/336 is 4.46%, which rounds to the printed 4. The same pattern occurs for anticholinergic agents (25/333 is 7.51%, printed 7; see C005). This suggests the olanzapine column may have used N=336 rather than the footnoted 333. It is a one-point rounding or denominator discrepancy for the authors to confirm.
Second check: another model agreed
Report a problem with this finding
Typo Figure 1, risperidone box reads "333 included in analysis" with a lowercase 'i'; the other four boxes read "Included in analysis".
- p. 6 Figure 1, bottom row, risperidone box
333 included in analysis
Judge's note: Seen on the page 6 image: the olanzapine, perphenazine, quetiapine and ziprasidone boxes have a capital I, and the risperidone box has lowercase. Cosmetic only.
Second check: none (judge only)
Report a problem with this finding
Typo Table 3 section header reads "Discontinuation of treatment owing to intolerability — no. %" without parentheses; the other headers use "no. (%)".
- p. 12 Table 3, section header
Discontinuation of treatment owing to intolerability — no. %
Judge's note: Seen on the page 12 image: this header prints 'no. %', while 'Adverse events — no. (%)' and the other headers have parentheses. Cosmetic only.
Second check: none (judge only)
Report a problem with this finding
Typo Table 3 (continued), medications added: row label "Cholestatin drugs" is not a standard drug-class name (presumably statins or cholesterol-lowering drugs).
- p. 13 Table 3 (Continued), Medications added
Cholestatin drugs
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that 'cholestatin' is not a recognised name for a class of lipid-lowering drugs. The label is printed as quoted on the page 13 image, and the paper does not define the term anywhere. The authors should confirm the intended label.
Second check: none (judge only)