The Ocular Hypertension Treatment Study: A Randomized Trial Determines That Topical Ocular Hypotensive Medication Delays or Prevents the Onset of Primary Open-Angle Glaucoma
Archives of Ophthalmology, 2002 · paper · Clinical trials
A randomized trial of 1636 people with raised eye pressure comparing topical pressure-lowering medication with observation. At 60 months, glaucoma had developed in 4.4% of the medication group versus 9.5% of the observation group, so treatment delayed or prevented onset.
3 findings
Substance The inactive totals by group in the Figure 4 risk table (Observation "Inactive ... 89", Medication "Inactive ... 84") are the reverse of Table 3 ("Inactive† 89 (10.9) 84 (10.2)" for medication and observation) and of Figure 1 (Observation n=84, Medication n=89).
- p. 7 Figure 4 at-risk table, Total column
Inactive 9 3 4 3 11 6 13 12 12 10 4 2 0 0 89
- p. 7 Table 3, Inactive row
89 (10.9) 84 (10.2)
- p. 4 Figure 1 flowchart
Lost to Follow-up (Inactive) (n=84)
- p. 7 Table 3, Inactive row
89 (10.9) 84 (10.2) 173 (10.6)
Only the authors can say what is right here.
Judge's note: Checked the page 7 and page 4 images. Each Figure 4 row adds up to its printed total, and each at-risk decrement equals POAG + deaths + inactive (for example 819-809 = 1+0+9 and 817-797 = 0+2+18), so the figure is internally consistent. Figure 1 completers (706 = 819-29-84 and 702 = 817-26-89) follow Table 3. Figure 4 defines its censoring as "withdrew before the end of the study", while Table 3 counts "missed their last 2 follow-up visits". These could be different quantities, but an exact swap of 89 and 84 points to a group-label exchange. The authors need to say which is right. Separately, Table 3's 84/819 = 10.26% is printed as 10.2.
Second check: another model agreed
Report a problem with this finding
Substance In Table 1 the Overall cell for age >60 to ≤70 reads "412 (25.6)". 412/1636 = 25.2%. The 25.6 matches the Observation column (210 (25.6)).
- p. 5 Table 1, Age >60 to ≤70, Overall column
412 (25.6)
Judge's note: Checked the Table 1 crop. The count 412 = 202+210 is correct. The other Overall age cells agree with N=1636 (578 → 35.3%, 529 → 32.3%, 117 → 7.2%). A gap of 0.4 points cannot come from rounding.
Second check: another model agreed
Report a problem with this finding
Reference or label Reference 38 prints "Arch Ophthalmol. 2000;18:1187-1194"; the volume should be 118.
- p. 13 Reference 38
Arch Ophthalmol. 2000;18:1187-1194
- p. 13 Reference 38
Arch Ophthalmol. 2000;18:1187-
Judge's note: The crop and the full page image both show '2000;18:1187-1194', so this is not an extraction artefact. The paper's own Arch Ophthalmol references fix the volume sequence: ref 31 1993;111, ref 7 1994;112, ref 36 1998;116, ref 33 1999;117, ref 12 2001;119, ref 41 2002;120, and the running footer 'ARCH OPHTHALMOL/VOL 120, JUNE 2002'. On that sequence the year 2000 is volume 118. Volume 18 cannot follow from the paper's own data, so this is an evident fix.
Second check: another model agreed
Report a problem with this finding
16 minor (typos and the like; most checked by the judge only)
Wording The text says "Recruitment was extended from 24 months to 30 months", but the window it gives, "Between February 28, 1994, and October 31, 1996, 3328 individuals were considered for study enrollment", covers about 32 months.
- p. 2 Results, Recruitment
Recruitment was extended from 24 months to 30 months
- p. 2 Results, Recruitment (continued)
Between February 28, 1994, and October 31, 1996,
Judge's note: Seen on page 2 (bottom right) continuing to page 3. February 28, 1994 to October 31, 1996 is 32 months. The dates are attached to people being "considered for study enrollment" (screening), so the 30 months may count from a different start point such as first randomization. The paper does not say this. The gap is small and could be legitimate, so this is an observation.
Second check: another model agreed
Report a problem with this finding
Wording The Comment section says "during a median follow-up period of 72 months", but the Results give no overall median, only "72 months for African American participants and 78 months for other participants".
- p. 7 Comment, paragraph beginning 'The OHTS demonstrated'
during a median follow-up period of 72 months
- p. 3 Results, Follow-up
The median duration of follow-up was 72 months for African American participants and 78 months for other participants
Only the authors can say what is right here.
Judge's note: Checked page 3 (Follow-up) and page 7 (Comment). The pooled median must lie between 72 and 78. It can equal 72 only if more than half of all participants have follow-up of 72 months or less, which with discrete visit timing is possible but not shown. The reviewer's working assumes F_AA(72) = 0.5, which is not required. 72 is the African American subgroup value, and the paper does not report an overall median. The authors should confirm the overall figure.
Second check: another model agreed
Report a problem with this finding
Wording The Comment section says "87% of the medication participants achieved this IOP target reduction in both eyes, and an additional 7% did so in one eye". The Results report these as shares of visits: "87% (7515 of 8621) and in one eye at 7% (613 of 8621) of the scheduled follow-up visits".
- p. 8 Comment, first column
87% of the medication participants achieved this IOP target reduction in both eyes, and an additional 7% did so in one eye
- p. 4 Results, IOP Reduction and Medication
The IOP goal was met in both eyes at 87% (7515 of 8621) and in one eye at 7% (613 of 8621) of the scheduled follow-up visits
- p. 4 Results, IOP Reduction and Medication
The IOP goal was met in both eyes at 87% (7515 of 8621) and in one eye at 7% (613 of 8621) of the scheduled follow-up visits completed by medication participants
Judge's note: Checked page 4 (Results) and page 8 (Comment). The figures are proportions of completed visits, not of the 817 participants. The Comment changes the unit.
Second check: another model agreed
Report a problem with this finding
Wording The Table 1 footnote says corneal thickness "Measurements were conducted after 1999, about 2 years after randomization of the last participant". Enrollment ended October 31, 1996, so measurements "after 1999" would be more than 3 years later.
- p. 5 Table 1 footnote
Measurements were conducted after 1999, about 2 years after randomization of the last participant.
Judge's note: Checked the footnote crop and the enrollment dates on pages 2-3. If "after 1999" means "from 1999 onward", the gap is about 2-2.5 years and the footnote holds. As printed the wording is ambiguous.
Second check: another model agreed
Report a problem with this finding
Typo The Executive/Steering Committee list has "Mae D. Gordon, PhD", but the byline and every other list have "Mae O. Gordon". Financial Disclosures also gives "Chris Johnson, MD" for the author listed as "Chris A. Johnson, PhD".
- p. 11 Committees, Executive/Steering Committee
Mae D. Gordon, PhD
- p. 1 Byline
Mae O. Gordon, PhD
- p. 12 Financial Disclosures
Chris Johnson, MD
- p. 11
Mae D. Gordon, PhD;
Judge's note: Seen in the page 11 crop ("Mae D. Gordon") and the page 12 crop ("Chris Johnson, MD"). The byline on page 1, the DSMC, the Endpoint Committee and the corresponding-author line all use Mae O. Gordon, and Chris A. Johnson is PhD elsewhere. Both are name/degree slips.
Second check: none (judge only)
Report a problem with this finding
Typo Financial disclosures give "Chris Johnson, MD: Humphrey Systems, consultant, grant support", but the byline and the committee and reading-center lists give "Chris A. Johnson, PhD".
- p. 12 Financial Disclosures
Chris Johnson, MD: Humphrey Systems, consultant
- p. 1 Byline
Chris A. Johnson, PhD
- p. 12 Financial Disclosures
Chris Johnson, MD:
Judge's note: Checked the crops. The page 1 byline reads "Chris A. Johnson, PhD". The page 12 disclosures read "Chris Johnson, MD: Humphrey Systems, consultant...". The text layer also lists "Chris A. Johnson, PhD" on the Steering Committee and as investigator for the Visual Field and SWAP reading centers (Portland). The Humphrey perimetry consultancy fits this investigator, so this is the same person with two different degrees. Cosmetic inconsistency.
Second check: none (judge only)
Report a problem with this finding
Typo Page 4 prints "63 (10.2%) of 614 other observation participants". 63/614 = 10.26%, which rounds to 10.3%.
- p. 4 POAG section, last paragraph
63 (10.2%) of 614 other observation participants
Judge's note: Checked on the page 4 image. The denominator 614 = 819-205 is correct. The other percentages in the same sentence round correctly: 22/614 = 3.6% and 26/205 = 12.7%. Only this one is a rounding slip.
Second check: none (judge only)
Report a problem with this finding
Typo In Table 1, African American in the Medication column reads "203 (25.0)". 203/817 = 24.8%.
- p. 5 Table 1, Race, African American, Medication column
203 (25.0)
Judge's note: Checked the Table 1 crop. The other race cells in the Medication column are correctly rounded (577 → 70.6%, 24 → 2.9%, 8 → 1.0%). The Observation cell, 205/819 = 25.0%, is correct.
Second check: none (judge only)
Report a problem with this finding
Typo In Table 1, African American in the Overall column reads "408 (25.0)". 408/1636 = 24.9%.
- p. 5 Table 1, Race, African American, Overall column
408 (25.0)
Judge's note: Checked the Table 1 crop. Every other cell is reported to one decimal and correctly rounded (for example 1137/1636 = 69.5%). 24.94 rounds to 24.9.
Second check: none (judge only)
Report a problem with this finding
Typo Page 5 prints cataract surgery as "6.4% (52 of 806)". 52/806 = 6.45%, which rounds to 6.5%.
- p. 5 Results, Safety paragraph
6.4% (52 of 806)
Judge's note: Checked on the page 5 image. The comparison figure, 35/813 = 4.3%, is correct. The difference is a rounding slip only.
Second check: none (judge only)
Report a problem with this finding
Typo The Figure 3 legend reads "Topical Carbonic Anydrase Inhibitors"; it should be "Anhydrase".
- p. 7 Figure 3 legend
Topical Carbonic Anydrase Inhibitors
Judge's note: The misspelling is visible in the crop of the Figure 3 legend on page 7.
Second check: none (judge only)
Report a problem with this finding
Typo The same Emory investigator is listed as "Anastasias Costarides, MD, PhD" in the clinical centers list and as "Anastasios Costarides, MD" in Financial Disclosures.
- p. 9 Clinical Centers, Emory University Eye Center
Anastasias Costarides, MD, PhD
- p. 12 Financial Disclosures
Anastasios Costarides, MD
Judge's note: Both spellings appear in the crops (pages 9 and 12). The internal inconsistency is certain. Which spelling is correct is outside knowledge; Anastasios is the usual form.
Second check: none (judge only)
Report a problem with this finding
Typo The author affiliation reads "Devers Eye Institute, Portland Ore (Dr Johnson)" with no comma before "Ore". Elsewhere it is "Portland, Ore".
- p. 10 Author affiliations
Devers Eye Institute, Portland Ore (Dr Johnson)
Judge's note: Seen in the page 10 crop. The clinical centers list has "Devers Eye Institute, Portland, Ore". Punctuation only.
Second check: none (judge only)
Report a problem with this finding
Typo The same donor company is named two ways: "Otsuka America Pharmaceutical Inc, Rockville, Md" in the acknowledgments and "Otsuka American Pharmaceutical, grant support" in the financial disclosures.
- p. 12 Financial Disclosures (Cioffi)
Otsuka American Pharmaceutical, grant support
- p. 11 Acknowledgment of drug donors
Otsuka America Phar-
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the company's legal name is Otsuka America Pharmaceutical, Inc. Within the paper, the inconsistency is checked. The page 11 crop shows "Otsuka America Phar-", which continues on page 12 as "maceutical Inc, Rockville, Md" (text layer). The page 12 disclosures crop shows "Otsuka American Pharmaceutical". The two forms differ, and the paper itself cannot settle which spelling is right.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 30 prints the coauthor as "Heijl A, Bengstsson B."; the usual spelling is Bengtsson.
- p. 13 Reference 30
Heijl A, Bengstsson B.
- p. 13 Reference 30
Bengstsson B
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the coauthor of Heijl's Graefes 2000 timolol trial is spelled 'Bengtsson'. The crop of page 13 shows 'Bengstsson' as printed. The name appears nowhere else in the paper, so the paper cannot confirm the correct spelling. Cosmetic misspelling.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 42 gives the journal as "Am J Oph-thamol. 1998;126:487-497"; it should read 'Am J Ophthalmol', which is how references 10, 35 and 43 spell it.
- p. 13 Reference 42
Am J Oph-thamol
- p. 13 Reference 42
Am J Oph-thamol. 1998;126:487-497
- p. 13 Reference 42
Am J Ophthamol. 1998;126:487-497.
Judge's note: The full page 13 image shows 'Am J Oph-' and 'thamol. 1998;126:487-497.' across a line break. The text layer agrees. Joining the two parts at the hyphen still gives 'Ophthamol', with the 'l' missing, whereas refs 10, 35 and 43 on the same page print 'Ophthalmol'. Cosmetic misspelling.
Second check: none (judge only)