Table 3.
Diagnosis Rates, Antibiotic Administration, and Waiting Times to See a Physician, Stratified by 2005 Pneumonia Antibiotic Timing Score*
| Antibiotic timing score quintile (2005) | |||||||
|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | P value for trend† | Adjusted P value for trend‡ | |
| Mean 2005 antibiotic timing score (range)§ | 60 (25 – 68) | 72 (69 – 76) | 80 (77 – 82) | 84 (83 – 86) | 91 (87 – 100) | -- | -- |
| ED diagnosis | % of visits with each diagnosis (95% CI) | ||||||
| Pneumonia | 10 (8 – 12) | 10 (8 – 12) | 10 (8 – 12) | 12 (10 – 14) | 13 (9 – 18) | .28 | .76 |
| Bronchitis | 20 (17 – 23) | 27 (23 – 31) | 24 (20 – 30) | 26 (23 – 30) | 31 (27 – 36) | <.001 | .09 |
| CHF | 8 (7 – 10) | 9 (7 – 11) | 7 (6 – 9) | 9 (6 – 12) | 4 (3 – 7) | .47 | .07 |
| Antibiotic usage among visits with… | % of visits receiving an antibiotic (95% CI) | ||||||
| …any ED diagnosis | 31 (26 – 35) | 37 (33 – 41) | 34 (31 – 38) | 37 (32 – 42) | 44 (41 – 48) | .004 | .17 |
| …an ED diagnosis of pneumonia | 73 (62 – 82) | 78 (68 – 86) | 81 (71 – 89) | 80 (68 – 88) | 81 (69 – 89) | .26 | .23 |
| …no ED diagnosis of pneumonia | 23 (19 – 28) | 29 (25 – 33) | 26 (22 – 30) | 28 (23 – 33) | 34 (29 – 39) | .04 | .39 |
| …an “inappropriate” ED diagnosis|| | 24 (19 – 30) | 24 (17 – 33) | 21 (15 – 28) | 26 (20 – 32) | 32 (25 – 41) | .84 | .68 |
| Mean waiting times to see a physician (minutes) | mean (95% CI) | ||||||
| Visits for respiratory symptoms | 66 (57 – 75) | 53 (44 – 62) | 40 (34 – 47) | 42 (32 – 51) | 38 (32 – 45) | <.001 | <.001 |
| Visits not for respiratory symptoms | 69 (62 – 75) | 57 (50 – 64) | 44 (40 – 49) | 44 (37 – 51) | 38 (32 – 44) | <.001 | <.001 |
| Difference, respiratory symptom vs. no respiratory symptom | 3 (–4 – 9) | 4 (–4 – 11) | 4 (–1 – 10) | 2 (–6 – 11) | –1 (–5 – 4) | .45 | .35 |
Abbreviations: ED, emergency department; CI, confidence interval; CHF, congestive heart failure.
Among reporting period (2004–2005) ED visits for respiratory symptoms to hospitals with ≥25 pneumonia antibiotic timing observations in 2005.
P values for test of trend reflect statistical significance of regression coefficient for continuous 2005 antibiotic timing score.
Adjusted P values reflect statistical significance of regression coefficients for continuous 2005 antibiotic timing score in models adjusting for patient- and hospital-level characteristics.
Score range in the lowest quintile has been rounded to shield hospital identity.
“Inappropriate” diagnoses are asthma, pulmonary edema, and CHF without concurrent diagnosis of pneumonia.