In the article beginning on page 4304 in the 22 November 2012 issue, there are errors in the Pediatric Hydroxyurea Phase 3 Clinical Trial data. These errors have two implications for the article’s findings. First, the rates of gastroenteritis are not significantly different between the hydroxyurea and placebo groups. In contrast, the rate of constipation is lower in the hydroxyurea group than the placebo group. Second, there is no significant difference in absolute neutrophil count for acute chest syndrome subjects between the hydroxyurea and placebo groups in Table 2. These errors only relate to adverse events and have no impact on the primary outcomes reported.
Table 2.
Characteristics of acute chest syndrome events
| Hydroxyurea, N = 96 | Placebo, N = 97 | P | |
|---|---|---|---|
| No. (%) | 8 (8.3) | 27 (27.8) | |
| Median age at event, mo (range) | 26 (15-30) | 27 (12-39) | .67 |
| Median visit week (range) | 46 (2-80) | 55 (0-99) | .22 |
| Associated symptoms | |||
| Fever | 7 (87.5) | 25 (92.6) | .55 |
| Dactylitis | 0 | 1 (3.7) | 1.00 |
| Pain | 4 (50) | 7 (25.9) | .23 |
| Absolute neutrophil count, < 0.5 × 109/L | 0 | 0 | 1.00 |
| Sepsis | 0 | 1 (3.7) | 1.00 |
| Pulmonary | |||
| > 1 lobe involved | 4 (50) | 12 (44.4) | 1.00 |
| Oxygen saturation at onset, % | 96 (91-99) | 94 (84-100) | .51 |
| Oxygen therapy | 3 (37.5) | 9 (33.3) | 1.00 |
| Management | |||
| Any hospitalization | 8 (100) | 27 (100) | 1.00 |
| Hospitalization > 7 d | 2 (25) | 2 (7.4) | .22 |
| Antibiotics | 8 (100) | 27 (100) | 1.00 |
| Transfusions | 2 (25) | 12 (44.4) | .43 |
| Simple | 1 | 12 | .14 |
| Exchange | 1 | 0 | .14 |
P values for categorical variables were derived from the Fisher exact test. P values for continuous variables were derived from the Wilcoxon rank-sum test.
In the “Absolute neutrophil count, < 0.5 × 109/L” row of Table 2 on page 4306, the number of subjects receiving a placebo should be 0 and the P value should be 1.00. The corrected Table 2 is shown below.
In the “Gastroenteritis” row of Table 3 on page 4307, all of the data are incorrect. The corrected Table 3 is shown below.
Table 3.
Etiology of febrile events
| Hydroxyurea, N = 96 | Placebo, N = 97 | Hazard ratio† | P† | |||||
|---|---|---|---|---|---|---|---|---|
| Events | Incidence rate* | Subjects, n (%) | Events | Incidence rate* | Subjects, n (%) | |||
| Fever | 352 | 186.2 | 86 (90) | 402 | 217.4 | 88 (91) | 0.89 | .5 |
| Aplastic crisis, parvovirus | 4 | 2.1 | 4 (4) | 4 | 2.2 | 4 (4) | 0.96 | 1 |
| Acute osteomyelitis | 0 | 0 | 0 | 2 | 1.1 | 2 (2) | — | .2 |
| Bacteremia/sepsis/meningitis | 3 | 1.6 | 2 (2) | 6 | 3.2 | 5 (5) | 0.40 | .3 |
| Viral syndrome | 41 | 21.7 | 30 (31) | 48 | 26.0 | 31 (32) | 0.92 | .7 |
| Gastroenteritis | 70 | 37.1 | 44 (46) | 69 | 37.3 | 45 (46) | 0.91 | .7 |
| Otitis media | 76 | 40.3 | 36 (38) | 99 | 53.5 | 45 (46) | 0.73 | .2 |
Per 100 patient-years.
The hazard ratio (hydroxyurea vs placebo) was generated from a Cox model and the P value from the log-rank life test to compare the time to first event between the two treatment groups.
In the “Infections” section on page 4307, the last two sentences of the first paragraph should be removed.
In the “Other adverse events” section on page 4308, the authors wish to add two new sentences immediately before the final sentence to reflect a significant result of new analysis. The sentences read, “Sixteen (17%) children in the hydroxyurea group and 39 (40%) children in the placebo group had constipation. The rate of constipation was 12.7 per 100 patient-years in the hydroxyurea group and 37.9 per 100 patient-years in the placebo group (HR 0.33; P < .01).”
In the “Discussion” section on page 4308, in the right column, the paragraph beginning “Unexpectedly” should be deleted.