Streptomycin Treatment of Pulmonary Tuberculosis: A Medical Research Council Investigation
BMJ, 1948 · paper · Biomedicine
Reports short-term results of a Medical Research Council controlled investigation of streptomycin in one type of pulmonary tuberculosis. The supplied text covers the committee, centres and rationale for controlled trials, without the results.
Checked: PDF from europepmc.org, fetched 4 Oct 2026. Page numbers are those of that PDF.
7 findings
Substance Treatment section: "One patient was treated with streptomycin for 6 months, 2 for 5½ months, 6 for 5 months, 5 for 4½ months; the remainder, 41 patients, were treated for 4 months" (14 and 41), but p. 14 has "five of 13 patients treated for more than four months, compared with 16 of 42 patients treated for four months only".
- p. 3 Treatment, right column, parenthesis
One patient was treated with streptomycin for 6 months, 2 for 5½ months, 6 for 5 months, 5 for 4½ months ; the remainder, 41 patients, were treated for 4 months.
- p. 14 Deterioration in S Cases, summary paragraph, left column
was seen in five of 13 patients treated for more than four months, compared with 16 of 42 patients treated for four months only
- p. 3 Treatment, parenthesis on duration
the remainder, 41 patients, were treated for 4 months
- p. 14 Paragraph beginning 'This analysis has shown'
five of 13 patients treated for more than four months, compared with 16 of 42 patients treated for four months only
Only the authors can say what is right here.
Judge's note: On the page 3 image the counts read 1, 2, 6, 5 and 41, with the fractions 5½ and 4½ legible. That gives 14 treated beyond four months and 41 for four months. The page 14 image prints 13 and 42. Both splits total 55, and 5 + 16 = 21 matches Table XI. The following sentence about one patient treated for five months does not explain moving a patient between groups, and I found no other explanation in the text. Only the authors can say which split is right.
Second check: another model agreed
Report a problem with this finding
Substance E.S.R. paragraph: "13 (68%) of 19 S patients were improved ... Eighteen (50%) of 36 S patients with E.S.R. over 50 were improved" gives 31 improved S patients, but Tables II and III give 38 S patients improved at six months.
- p. 4 Sedimentation Rate (E.S.R.) paragraph, right column
13 (68%) of 19 S patients were improved
- p. 4 Sedimentation Rate (E.S.R.) paragraph, right column
Eighteen (50%) of 36 S patients with E.S.R. over 50 were improved
- p. 3 Table II, Streptomycin Group
28 51%
- p. 3 Table II
Moderate or slight improvement 10 18%
Only the authors can say what is right here.
Judge's note: Read the E.S.R. paragraph and Table III on the page 4 image and Tables I and II on the page 3 image. The two E.S.R. strata (19 and 36) cover all 55 S patients, yet the improved counts total 31. Table II gives 28 + 10 = 38 improved and Table III gives 3 + 9 + 13 + 13 = 38. The temperature analysis reconciles (26 of 39 at 100°F or over). The C figures reconcile too: 9 + 7 = 16 against 17 improved, with one E.S.R. not done per the Table I footnote. Each printed percentage matches its own count, so the paper does not show which S count is wrong. No alternative definition of "improved" is given.
Second check: another model agreed
Report a problem with this finding
Substance Summary: "Apart from those who died, deterioration was seen in 18% of S cases and 34% of C cases"; Table II gives S deterioration as 5 (9%) + 6 (11%) = 11 of 55 = 20%.
- p. 18 Summary, paragraph beginning 'At the end of six months'
deterioration was seen in 18% of S cases and 34% of C cases
- p. 3 Table II, Streptomycin Group, deterioration rows
5 9%
- p. 3 Table II, Streptomycin Group, considerable deterioration
6 11%
- p. 3 Table II
Moderate or slight deterioration 5 9% ... Considerable deterioration 6 11%
Judge's note: Read the Summary on the page 18 image and Tables II and IX on the page 3 and page 10 images. S deterioration excluding deaths is 5 + 6 = 11 of 55, which is 20%. The C figure agrees with the table (12 + 6 = 18 of 52, printed 23% + 11% = 34%). The other Summary percentages (7%, 27%, 51%, 8%, 18%, 25%) match Table II. The 18% appears to repeat the S slight-or-moderate-improvement figure from the sentence before. No other time point or denominator gives 18% (11/51 is 21.6%).
Second check: another model agreed
Report a problem with this finding
Substance Score table: "S cases .. 71 124 98"; applying the stated scoring to the S third-month rows of Table XIII gives 121, not 124.
- p. 15 Score table, S cases, 'During Third Month'
124
- p. 14 Table XIII, S Cases, Results in Third Month
40 0 16 12 10 2
- p. 15 numerical score table, S cases, During Third Month
S cases .. 71 124 98
- p. 14 Table XIII, S cases, results in third month
Smear strongly positive 40 0 16 12 10 2
Only the authors can say what is right here.
Judge's note: Read Table XIII on the page 14 image and the score table on the page 15 image. S third-month column totals are 18 strongly positive, 15 weakly positive, 11 culture positive, 10 negative and 0 deaths (54 cases). With the stated scoring (1, 2, 3, 4, 0) that is 121. The same method reproduces the other five printed scores exactly: S 71 and 98, C 75, 69 and 62. The table cells agree with the text (10 S negative; 12 + 10 + 1 = 23 at a lower level), which points to the score, but only the authors can say whether a cell or the score is the misprint.
Second check: another model agreed
Report a problem with this finding
Substance Severity breakdown "0/2 where resistance was not more than 10 times H37Rv, 3/7 ..." and "0/5 where resistance was not more than 10 times H37Rv, 2/14 ..." gives 7 and 21 patients, against "Of six patients ... did not show resistance greater than 10 times" and "Twenty-two patients developed streptomycin resistance over 10 and less than 1,000 times".
- p. 16 Results Related to Resistance Development, left column
Of six patients from whom strains isolated did not show resistance greater than 10 times that of the standard H37Rv, all had improved
- p. 16 right column, very ill patients
0/2 where resistance was not more than 10 times H37Rv, 3/7 where it was more than 10
- p. 16 right column, other patients
0/5 where resistance was not more than 10 times H37Rv, 2/14 where it was more than
- p. 16 left column
Twenty-two patients developed streptomycin resistance over 10 and less than 1,000 times
Only the authors can say what is right here.
Judge's note: Read both columns of the page 16 image. The stratified denominators give 2 + 5 = 7 in the lowest band and 7 + 14 = 21 in the middle band; the same page states 6 and 22. Page 15 supports 6 and 22 (six cases not above 10 times; 5 + 17 = 22 between 32 and 256 times). Everything else agrees: the top band is 6 + 7 = 13, the total is 41, the severity split is 15 and 26 ("six of the 13" plus "nine of the 28"), and the deteriorated counts are 3 + 2 = 5 and 6 + 2 = 8. One patient among the "other patients" appears to be in the wrong band. Chart VI seems to show six points at or below 10 times (two squares, four dots), which would suggest 0/4 and 2/15, but that reading is from a low-resolution chart.
Second check: another model agreed
Report a problem with this finding
Reference or label Plate caption "FIG. 13.—Case 40. Jan. 30, 1947." should read Case 46: the text cites "considerable improvement (see Figs. 13 and 14)" for Case 46, and Figs. 14–16 are captioned Case 46.
- p. 8 Plate, caption of Fig. 13
FIG. 13.—Case 40. Jan. 30, 1947.
- p. 13 Case 46 history, right column
considerable improvement (see Figs. 13 and 14)
- p. 8 Figs. 14-16 captions
FIG. 14.—Case 46. March 26, 1947.
Judge's note: On the page 8 plate the Fig. 13 caption reads "Case 40", while Figs. 14, 15 and 16 read "Case 46". The film in Fig. 13 carries the number 460, and the series runs 460, 462, 464, 466. The Case 46 history on page 13 cites Figs. 13 and 14 for the early improvement and Figs. 15 and 16 for the later deterioration. Case 40 is a different patient with no figure reference. Caveat: the caption is small rotated type read from the full-page image only, as no crop was supplied.
Second check: another model agreed
Report a problem with this finding
Reference or label Chart IV: the first S bar ("ADMISSION TO END OF 2ND MONTH") is drawn at roughly 86–87%, whereas Table X gives 8 + 34 = 42 of 55 and the text says "At two months 76% of S patients showed radiological improvement".
- p. 10 Chart IV, first S bar
ADMISSION TO END OF 2ND MONTH
- p. 10 Text, right column
At two months 76% of S patients showed radiological improvement
Only the authors can say what is right here.
Judge's note: Chart IV is on PDF page 11, not page 10 as the cluster says. On that image the first S bar tops out between the 80 and 90 gridlines, near 87. The other bars read at their tabulated values: S about 65, 34 and 69; C about 6, 18, 19 and 33. The considerable-improvement segment of the first bar is at about 14, as tabulated. Table X (page 10 crop) gives 14% + 62% = 76%. The bar appears plotted about 10 points too high. This is read from a page image, so the size of the gap is approximate.
Second check: another model agreed
Report a problem with this finding
4 minor (typos and the like; most checked by the judge only)
Wording Menstruation paragraph: "In 10 of 32 female S patients and in 12 of 31 female C patients", whereas the S group is stated to contain "22 men and 33 women" (and "two of 33 female S patients" on p. 4).
- p. 3 Condition on Admission, last sentence before X-ray Classification
There were 22 men and 33 women in the S group
- p. 9 Menstruation paragraph
In 10 of 32 female S patients and in 12 of 31 female C patients
- p. 9 Menstruation paragraph
In 10 of 32 female S patients
- p. 3 Condition on Admission
22 men and 33 women in the S group
Only the authors can say what is right here.
Judge's note: Checked pages 3, 4 and 9. The S menstruation categories (10, 9, 11, 2) sum to 32, so the paragraph is internally consistent but one S woman short of the 33 stated elsewhere. The C figures (12, 12, 7) sum to 31 and match. The paper gives no missing-data note for menstruation, unlike the footnotes on Tables I, VI and VII. Most likely one woman had no data, but that is not stated.
Second check: another model agreed
Report a problem with this finding
Wording "in the C group the number of patients with a very high sedimentation rate (over 50) was never reduced", but Table VII gives C 51+ plus deaths as 29 on admission and 14 + 14 = 28 at six months.
- p. 9 Text below Table VII
the number of patients with a very high sedimentation rate (over 50) was never reduced
- p. 9 Table VII, C group, 51+ and Deaths rows
29 37 23 14
Judge's note: Read Table VII on the page 9 image. C group, E.S.R. over 50 plus deaths: 29, then 37 + 2 = 39, then 23 + 10 = 33, then 14 + 14 = 28. The six-month figure is one below admission, so "never reduced" is slightly loose. The table footnote says totals differ because results were not available in all cases (C total 51 on admission, 50 later), so the one-patient difference may reflect a missing result.
Second check: another model agreed
Report a problem with this finding
Typo Text below Table V: "(temperature 99–100.9° F.; 37.2–38.85° C.)"; the Table V row label for the same band reads "(37·2–38·25° C.)".
- p. 9 Text below Table V, left column
37.2–38.85° C.
- p. 9 Table V, second row label
(37·2–38·25° C.)
- p. 9 text below Table V, 'less acutely febrile patients'
temperature 99-100.9° F.; 37.2-38.85° C.
- p. 9 text after Table V
(temperature 99-100.9° F.; 37.2-38.85° C.)
Judge's note: The page 9 image shows 38.85 in the running text and 38.25 in the Table V stub; Tables I and III also give 38.25 for this band. Converting 100.9°F gives about 38.3°C, so 38.85 is a misprint for 38.25.
Second check: none (judge only)
Report a problem with this finding
Typo "Of 35 cases showing resistance over 10 times H37, this resistance emerged in five cases in the first month" — the standard strain is written H37Rv everywhere else.
- p. 15 Time of Emergence of Streptomycin Resistance
resistance over 10 times H37, this resistance emerged
Judge's note: The page 15 image prints "H37" in this sentence, while the same page and Chart V caption use "H37Rv". It is a truncated strain name; the meaning is unambiguous and it could be shorthand.
Second check: none (judge only)