10-Year Outcomes after Monitoring, Surgery, or Radiotherapy for Localized Prostate Cancer
2016 · paper · Clinical trials
A trial randomizing 1643 men with PSA-detected localized prostate cancer to active monitoring, surgery or radiotherapy. At about 10 years, prostate-cancer deaths were low and similar across groups, though surgery and radiotherapy reduced disease progression and metastases compared with monitoring.
0 findings
None that passed a second check.
5 minor (typos and the like; most checked by the judge only)
Wording Methods, p. 3: "androgen-deprivation therapy was offered when PSA levels reached 20 ng per milliliter or less, if indicated". The direction of the trigger threshold is unclear as printed.
- p. 3 Treatment Procedures and Clinical Management, right column
apy was offered when PSA levels reached 20 ng per milliliter or less, if indicated
- p. 3 Treatment procedures
androgen-deprivation ther-apy was offered when PSA levels reached 20 ng per milliliter or less
- p. 3 same paragraph
Imaging of the skeleton was recommended if the PSA level reached 10 ng per milliliter.
- p. 3 Methods, Treatment Procedures
therapy was offered when PSA levels reached 20 ng per milliliter or less, if indicated
Only the authors can say what is right here.
Judge's note: Checked the crop and page 3. The words 'reached 20 ng per milliliter or less, if indicated' are printed as quoted. The next sentence recommends skeletal imaging at PSA 10 ng/ml, and the other triggers in the paragraph are stated as 'or higher'. The intended meaning may be 'at 20 ng/ml, or at lower levels if clinically indicated', so it is not certain that 'or more' is what the authors meant. Either way the printed wording is ambiguous or illogical, and only the authors can say which rule applied.
Second check: another model agreed
Report a problem with this finding
Wording The final Discussion paragraph says "44% of the patients who were assigned to active monitoring did not receive radical treatment". The Results report 291 of 545 radically treated (Kaplan–Meier estimate, 54.8%), which leaves about 45–47% untreated.
- p. 9 Discussion, final paragraph
44% of the patients who were assigned to active monitoring did not receive radical treatment
- p. 4 Results, Adherence
291 had received a radical treatment by the end of November 2015 (Kaplan–Meier estimate, 54.8%
- p. 4 Results, Adherence to Assigned Treatment
Meier estimate, 54.8%; 95% confidence interval
- p. 9 Discussion, first lines
almost half the men received no intervention during the 10-year follow-up
Only the authors can say what is right here.
Judge's note: Viewed page 9 (the 44% sentence and 'almost half') and page 4 (291/545; KM 54.8%, 95% CI 50.4 to 59.3). Neither complement rounds to 44%. The 44% is cited to ref 5 (Donovan et al., the companion patient-reported outcomes paper), which may use a different cohort or time point, but this paper does not say so. The authors should say where the figure comes from.
Second check: another model agreed
Report a problem with this finding
Wording The all-cause mortality P value is credited to two different tests. The Results say "P = 0.87 by likelihood-ratio test" and cite Table 1. The Table 1 footnote says "P values were calculated with the use of a log-rank test ... The planned adjusted analysis was not possible".
- p. 5 Results, Prostate-Cancer–Specific and All-Cause Mortality
(P=0.87 by likelihood-ratio test)
- p. 6 Table 1 footnote *
P values were calculated with the use of a log-rank test of the null hypothesis of no difference in effectiveness across the three treatments.
- p. 5 Results, Mortality
P = 0.87 by likelihood-ratio test
- p. 6 Table 1 footnote *
P values were calculated with the use of a log-rank test
Only the authors can say what is right here.
Judge's note: Viewed the Table 1 crop on p. 6: the all-cause row prints 0.87, and the footnote attributes every P value in the table to a log-rank test. The Results on p. 5 give the same 0.87 for the same three-group comparison, cite Table 1, and name a likelihood-ratio test. The parallel prostate-cancer P value (0.48) is called log-rank in the text. The Statistical Analysis section mentions only adjusted Cox models. This looks like one quantity labelled two ways. The authors should say which test produced 0.87; Table S3, which the text also cites, is not among the files.
Second check: another model agreed
Report a problem with this finding
Typo Discussion, p. 7: "receptor activator of nuclear factor-κβ ligand" is printed with a Greek beta where the standard name uses a Latin B (κB).
- p. 7 Discussion, right column below Figure 3
receptor activator of nuclear factor-κβ ligand
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the standard name is nuclear factor-κB, with a Latin B. The page 7 crop shows an italic Greek β after κ, so this is not an extraction artefact.
Second check: none (judge only)
Report a problem with this finding
Typo Discussion, p. 7: "bone-targeted agents (e.g., radionuclides, inhibitors of ... [RANKL], and chemotherapy), and immunotherapy". Because of where the parenthesis closes, chemotherapy is listed as an example of a bone-targeted agent.
- p. 7 Discussion, second paragraph
bone-targeted agents (e.g., radionuclides, inhibitors of receptor activator of nuclear factor-κβ ligand [RANKL], and chemotherapy), and immunotherapy
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: chemotherapy is not a bone-targeted agent. The page 7 crop shows that the closing parenthesis comes after 'chemotherapy', so the text reads as quoted. This is most likely a punctuation slip in a list of treatment classes.
Second check: none (judge only)