Robot-Assisted Radical Cystectomy versus Open Radical Cystectomy in Patients with Bladder Cancer (RAZOR)
2018 · paper · Clinical trials
A randomized noninferiority trial of 350 bladder cancer patients, comparing robot-assisted with open bladder removal. Two-year progression-free survival was 72.3% versus 71.6%, meeting the noninferiority criterion, with similar adverse event rates.
6 findings
Substance Text "40 (29%) of 150 patients in the robotic cystectomy group and 27 (18·5%) of 152 patients in the open cystectomy group" uses denominators that do not give the percentages; Table 3 gives 40/139 (29%) and 27/146 (18%).
- p. 7 Results text, second column
40 (29%) of 150 patients in the robotic cystectomy group and 27 (18·5%) of 152 patients
- p. 7 Table 3, Hospital stay ≤5 days
40/139 (29%)
- p. 7 Results text
40 (29%) of 150 patients in the
- p. 7 Results text vs Table 3
27 (18·5%) of 152 patients
Judge's note: Checked on the page 7 image. The percentages in the text only reproduce with the Table 3 denominators of 139 and 146. 150 and 152 are the per-protocol arm sizes. The text's 18·5% also differs from Table 3's 18%.
Second check: another model agreed, code check
Report a problem with this finding
Substance Table 3 open-cystectomy T stage counts (T0 "31 (20%)", 25, 15, 33, 32, 14, 3) sum to 153, although n=152; T0=31 also differs from 'No residual disease' 30 (20%).
- p. 7 Table 3, T stage, open cystectomy column
31 (20%)
- p. 7 Table 3, T stage
T0 22 (15%) 31 (20%)
Only the authors can say what is right here.
Judge's note: Table 3 checked on the page 7 image. Every other open-column block adds up to 152: complications 47+20+51+28+2+4, histopathology 116+4+29+3, N stage 7+121+8+13+3, pathological stage 30+24+15+29+29+25. In the robotic column, T0 (22) equals no residual disease (22). The footnote explains nothing. Likely T0 should be 30, but only the authors can confirm.
Second check: another model agreed
Report a problem with this finding
Substance Results text: "40 (29%) of 150 patients in the robotic cystectomy group and 27 (18·5%) of 152 patients in the open cystectomy group". The percentages use the Table 3 evaluable denominators (40/139, 27/146), not the 150 and 152 the text states.
- p. 7
27 (18·5%) of 152
Judge's note: The page 7 text gives 27 (18·5%) of 152, but 27/152 = 17·8%. Table 3 prints Hospital stay ≤5 days as 40/139 (29%) and 27/146 (18%). 27/146 = 18·49% and 40/139 = 28·8%, so the percentages were computed on the evaluable denominators while the text names the full arms (40/150 would be 26·7%). The text also says 'less than 5 days' where the table says '≤5 days'. Nothing in the paper explains the mismatch.
Second check: another model agreed, code check
Report a problem with this finding
Reference or label Results text says "major complications (grades III–IV) was also similar between the groups (table 3)", but the only major-complication comparison in Table 3 is "Grades III–V vs 0–II" (33 [22%] vs 34 [22%]).
- p. 8 Results text, left column
major complications (grades III–IV) was also similar
- p. 7 Table 3
Grades III–V vs 0–II
- p. 2 Research in context
major complications (Clavien-Dindo grade ≥3)
- p. 8 Results, paragraph after table 3
major complications (grades III–IV)
Judge's note: Table 3 on page 7 gives III 29/28, IV 0/2, V 4/4, and the row 'Grades III–V vs 0–II' 33 vs 34, p=0·94. The counts include grade V deaths (29+0+4 and 28+2+4). Research in context on page 2 uses grade ≥3. No grade III–IV analysis is reported anywhere, so the text's range is a mislabel.
Second check: another model agreed
Report a problem with this finding
Reference or label Discussion: "refutes retrospective study comparisons4,8–10,16,18 reporting significantly lower complication rates for robotic surgery". Refs 8, 16 and 18 are randomised trials, and ref 4 is a morbidity-reporting study.
- p. 10 Discussion, left column
refutes retrospective study comparisons4,8–10,16,18 reporting significantly lower complication rates
- p. 11 Reference list, refs 8, 16, 18
Prospective randomized controlled trial of robotic versus open
- p. 10 Discussion, paragraph on surrogates of surgical quality
refutes retrospective study comparisons4,8–10,16,18 reporting significantly lower complication rates for robotic surgery
- p. 2 Research in context, Evidence before this study
no difference in complication rates, but better perioperative outcomes
Only the authors can say what is right here.
Judge's note: Reference list: ref 8 is Nix et al., 'Prospective randomized controlled trial of robotic versus open radical cystectomy'. Ref 16 is Parekh et al., 'pilot prospective randomized clinical trial'. Ref 18 is Bochner et al., 'a randomized clinical trial'. Ref 4 is Shabsigh et al., 'Defining early morbidity of radical cystectomy'. Research in context (page 2) says the pilot randomised studies found 'no difference in complication rates'. The same Discussion sentence says the trial's result is concordant with single-centre randomised studies. The citation group appears misplaced, or meant to cover the whole sentence; only the authors can say which.
Second check: another model agreed
Report a problem with this finding
Reference or label Text: intraoperative and postoperative transfusion lower "(p=0·0002 and p=0·0089, respectively; table 3)", but Table 3 gives p<0·0001 for intraoperative transfusion; 0·0002 is the perioperative transfusion p value.
- p. 7 Results, first paragraph
(p=0·0002 and p=0·0089, respectively; table 3)
- p. 7 Table 3, Intraoperative transfusion row
18/139 (13%)
- p. 7 Table 3, intraoperative transfusion
<0·0001
- p. 7 Table 3, Intraoperative transfusion row
–20·8 (-30·6 to –11·2)
Judge's note: Table 3 on the page 7 image gives perioperative 35/143 vs 65/143, p=0·0002; intraoperative 18/139 vs 46/136, p<0·0001; postoperative p=0·0089. The text gives the intraoperative outcome the perioperative p value. The fix is p<0·0001, or naming perioperative transfusion. Both are below the threshold either way.
Second check: another model agreed
Report a problem with this finding
13 minor (typos and the like; most checked by the judge only)
Wording Research in context: "We searched PubMed for studies published between Jan 1, 2011, and Aug 10, 2017, without date or language restrictions" gives a publication-date window and also says there was no date restriction.
- p. 2 Research in context, Evidence before this study
between Jan 1, 2011, and Aug 10, 2017, without date or language restrictions
- p. 2 Research in context
published between Jan 1, 2011, and Aug 10, 2017, without date or language restrictions
Only the authors can say what is right here.
Judge's note: Checked on the page 2 image (crop C001-a/b). The sentence gives a date window and also says there were no date restrictions, so it contradicts itself. Only the authors can say which one is meant (for example, whether the dates are when the search was run).
Second check: another model agreed
Report a problem with this finding
Wording Methods describe "17 questions assessing bladder cancer-specific QoL, which are scored on a 5-point Likert scale", but the Table 4 footnote gives a range of "0–60 for FACT-BL-Cys", which fits 15 items scored 0–4, not 17.
- p. 3 Methods, Procedures
17 questions assessing bladder cancer-specific QoL, which are scored on a 5-point Likert scale
- p. 8 Table 4 footnote
0–60 for FACT-BL-Cys
- p. 3 Methods, Procedures
17 questions
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: FACT items are scored 0–4. Page 3 Methods and the page 8 Table 4 footnote were checked on the images. The paper does not say that any of the 17 items are unscored. The derived ranges (116, 168) follow from 60. The authors should state the number of scored items.
Second check: another model agreed
Report a problem with this finding
Wording Figure 1 boxes "159 had robotic cystectomy" and "153 had open cystectomy" include patients who received the other procedure ("4 underwent open cystectomy*", "1 underwent robotic cystectomy*"), so these boxes count patients who had surgery, not the named procedure.
- p. 4 Figure 1 trial profile
159 had robotic cystectomy
- p. 4 Figure 1 trial profile
4 underwent open cystectomy*
Judge's note: Figure 1 checked on the page 4 image. Each box feeds the mITT set (159 and 153), but its label names the assigned procedure even though 4 robotic-arm and 1 open-arm patients had the wrong procedure. The label does not match what is counted.
Second check: another model agreed
Report a problem with this finding
Wording The Discussion says chemotherapy and extended lymph node dissection "were equally balanced between groups", but Table 1 shows neoadjuvant chemotherapy in 41 (27%) vs 55 (36%) and adjuvant chemotherapy in 25 (17%) vs 17 (11%).
- p. 11 Discussion, limitations paragraph (continued from p 2534)
these treatments were equally balanced between groups
- p. 4 Table 1, chemotherapy rows
41 (27%)
- p. 4 Table 1, neoadjuvant chemotherapy, open group
55 (36%)
Judge's note: Table 1 on page 4 gives perioperative chemotherapy 62 (41%) vs 70 (46%), neoadjuvant 41 (27%) vs 55 (36%), and adjuvant 25 (17%) vs 17 (11%). Table 3 gives extended lymph node dissection 51% vs 55%, p=0·46. 'Equally balanced' overstates the similarity for neoadjuvant chemotherapy, though the authors may mean 'not significantly different'. This is a wording point.
Second check: another model agreed
Report a problem with this finding
Wording Text says patients "stayed in hospital for less than 5 days after surgery (p=0·0407; table 3)", while Table 3 row is "Hospital stay ≤5 days" with the same counts and p.
- p. 7 Results text
stayed in hospital for less than 5 days after surgery (p=0·0407; table 3)
- p. 7 Table 3
Hospital stay ≤5 days
Judge's note: Checked on the page 7 image. The counts (40 and 27) and p=0·0407 are the same, but the text uses a strict cut-off and the table an inclusive one. The text should say '5 days or less'.
Second check: another model agreed
Report a problem with this finding
Wording Table 3 Grades III–V vs 0–II difference "–0·4 (–9·0 to 9·8)": the interval is centred on +0·4, not on the printed estimate –0·4, unlike the other difference CIs in the table.
- p. 7 Table 3, Grades III–V vs 0–II row
–0·4 (–9·0 to 9·8)
Only the authors can say what is right here.
Judge's note: Checked on the page 7 image. 33/150 − 34/152 = −0·37 pp, and the Wald SE gives about −9·7 to 9·0. The same Wald method reproduces the other rows: grades I–V −1·75 (−12·25 to 8·75), printed −1·8 (−12·3 to 8·8); perioperative transfusion −21·0 (−31·8 to −10·2). The printed interval's midpoint is +0·4, consistent with a sign flip. The estimate still lies inside the interval, so this is a query on how the interval was computed.
Second check: another model agreed
Report a problem with this finding
Typo Affiliation reads "Department of Urology, Oklahoma University of Oklahoma, Norman, OK, USA" with 'Oklahoma' duplicated.
- p. 2 Affiliations, margin
Department of Urology, Oklahoma University of Oklahoma
Judge's note: The margin affiliation on the page 2 image reads 'of Urology, Oklahoma / University of Oklahoma,'. The extra 'Oklahoma' is a copyediting slip.
Second check: none (judge only)
Report a problem with this finding
Typo "Written informed consent was obtained from all patients" lacks a terminal full stop.
- p. 3 Patients
Written informed consent was obtained from all patients
Judge's note: The page 3 image shows no full stop at the end of the Patients paragraph.
Second check: none (judge only)
Report a problem with this finding
Typo The Table 4 footnote expands FACT as "Functional Assessment of Therapy" (FACT-VCI, FACT BL-Cys, FACT-G), but the Methods give "Functional Assessment of Cancer Therapy (FACT)".
- p. 8 Table 4 footnote
FACT-VCI=Functional Assessment of Therapy-Vanderbilt Cystectomy Index
- p. 3 Methods, Procedures
Functional Assessment of Cancer Therapy (FACT)-
Judge's note: The page 8 footnote reads 'FACT-VCI=Functional Assessment of Therapy-Vanderbilt Cystectomy Index. FACT BL-Cys=Functional Assessment of Therapy for patients with Bladder Cancer... FACT-G=Functional Assessment of Therapy-General form.' The page 3 Methods read 'Functional Assessment of Cancer Therapy (FACT)'. 'Cancer' is missing from all three expansions in the footnote.
Second check: none (judge only)
Report a problem with this finding
Typo "(difference 0·7% [95% CI –9·6 to 10·9; p=0·90; pnon-inferiority=0·001)": the square bracket opened before '95% CI' is never closed.
- p. 6 Results, right column
(difference 0·7% [95% CI –9·6 to 10·9; p=0·90;
- p. 6 Results, para 2
difference 0·7% [95% CI –9·6 to 10·9; p=0·90; pnon-inferiority=0·001)
- p. 6 Results, paragraph 2
p=0·90; pnon-inferiority=0·001
- p. 6 Figure 2A
p=0·756
Judge's note: The page 6 image confirms the unclosed square bracket. The p-value part of this cluster (text p=0·90 vs Figure 2A p=0·756) is not a defect. The Methods (statistical analysis) say groups were compared by the log-rank test (the figure p), while the 2-year PFS difference and its CI and non-inferiority test were 'derived using methods accounting for censored observations'. The text p goes with the 2-year difference and its CI, so the two p values test different quantities.
Second check: none (judge only)
Report a problem with this finding
Typo "ten (7%) of 150 patients in the robotic cystectomy were censored ... 14 (9%) of 152 patients in the open cystectomy were censored" is missing the word 'group' twice.
- p. 6 Results, right column
150 patients in the robotic cystectomy were censored
Judge's note: Checked on the page 6 image, right column.
Second check: none (judge only)
Report a problem with this finding
Typo Table 3, Intraoperative transfusion: difference printed "–20·8 (-30·6 to –11·2)", but 18/139 vs 46/136 gives –20·87 (rounds to –20·9), and the CI midpoint is –20·9.
- p. 7 Table 3, Intraoperative transfusion row
–20·8 (-30·6 to –11·2)
Judge's note: I checked the row on the page 7 crop: 18/139 (13%) vs 46/136 (34%), difference –20·8. The unrounded difference is –20·87, which rounds to –20·9. The other rows match unrounded arithmetic (–21·0, –15·0, 10·3). However, subtracting the one-decimal percentages (13·0 – 33·8) gives –20·8, so the printed value may come from rounding the intermediate figures. The discrepancy is 0·1, so I treat it as a minor consistency point.
Second check: none (judge only)
Report a problem with this finding
Typo Missing word: "two (1%) of 152 patients in the open cystectomy. No" should read "in the open cystectomy group".
- p. 8 Results, left column
two (1%) of 152 patients in the open cystectomy. No
Judge's note: Checked on page 8, left column. The parallel phrase earlier in the same sentence reads 'robotic cystectomy group'.
Second check: none (judge only)