Table 3.
Characteristics associated with high adherence to LTOT in the short-, medium-, and long-term by oxygen prescription group
| Continuous Group | Intermittent Group | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Odds ratio | 95% CI | P | Odds ratio | 95% CI | P | ||||
| Short term (1st 30 days; N=212) | Short term (1st 30 days; N=145) | ||||||||
| Age at trial entry (per year) | 1.00 | (0.96, 1.05) | 0.99 | Age at trial entry (per year) | 1.06 | (0.99, 1.12) | 0.09 | ||
| White race (vs. other) | 0.57 | (0.22, 1.45) | 0.24 | White race (vs. other) | 0.73 | (0.23, 2.36) | 0.60 | ||
| Female (vs. male) | 0.85 | (0.42, 1.72) | 0.66 | Female (vs. male) | 2.08 | (0.72, 5.99) | 0.18 | ||
| Self-reported rating of readiness to use oxygen at randomization (per unit) | 1.21 | (1.05, 1.40) | 0.008 | Self-reported rating of readiness to use oxygen at randomization (per unit) | 1.94 | (1.45, 2.59) | <0.001 | ||
| Exacerbation in 3 mos prior to trial entry (yes vs. no) | 2.37 | (0.97, 5.78) | 0.06 | Exacerbation in 3 mos prior to trial entry (yes vs. no) | 0.37 | (0.12, 1.11) | 0.08 | ||
| Ever participated in pulmonary rehabilitation as of trial entry (yes vs. no) | 3.02 | (1.29, 7.05) | 0.01 | Ever diagnosed with cardiovascular disease as of trial entry (yes vs. no) | 0.22 | (0.08, 0.60) | 0.003 | ||
| Self-reported rating of confidence in using oxygen at randomization (per unit) | 1.24 | (1.03, 1.50) | 0.03 | MMRC dyspnea score at trial entry (per unit) | 0.58 | (0.32, 1.04) | 0.07 | ||
| Current tobacco smoker at trial entry (yes vs. no) | 0.49 | (0.25, 0.96) | 0.04 | SGRQ total score at trial entry (per unit) | 1.02 | (0.99, 1.05) | 0.16 | ||
| Model goodness of fit: Hosmer-Lemeshow χ2 (df=3)= 4.30 withP=0.23, indicating good fit. | Model goodness of fit: Hosmer-Lemeshow χ2 (df=3)= 3.84, P=0.28, indicating good fit. | ||||||||
| Medium term (months 9–12; N=208) | Medium term (months 9–12; N=139) | ||||||||
| Age at trial entry (per year) | 0.96 | (0.91, 1.01) | 0.10 | Age at trial entry (per year) | 1.11 | (1.03, 1.20) | 0.007 | ||
| White race (yes vs. no) | 0.94 | (0.33, 2.70) | 0.90 | White race (vs. other) | 0.68 | (0.15, 2.97) | 0.60 | ||
| Female (vs. male) | 0.35 | (0.15, 0.86) | 0.02 | Female (vs. male) | 0.24 | (0.07, 0.79) | 0.02 | ||
| High adherence in the 1st 30 days since starting oxygen (yes vs. no) | 15.85 | (7.24, 34.73) | <0.001 | High adherence in the 1st 30 days since starting oxygen (yes vs. no) | 5.11 | (1.63, 15.97) | 0.005 | ||
| Self-reported rating of importance of using oxygen at randomization (per unit) | 1.23 | (1.05, 1.44) | 0.01 | Self-reported rating of importance of using oxygen at randomization (per unit) | 1.38 | (1.10, 1.73) | 0.005 | ||
| Exacerbation in 1st 8 months since starting oxygen (yes vs. no) | 2.32 | (0.99, 5.43) | 0.05 | Wheeze or whistling in chest at trial entry (yes vs. no) | 6.15 | (2.03, 18.68) | 0.001 | ||
| Post BD FEV1 at trial entry (per L) | 0.45 | (0.18, 1.10) | 0.08 | 6-minute walk distance at trial entry (per 10 ft) | 0.97 | (0.95, 0.99) | 0.002 | ||
| Ever used home oxygen before trial entry (yes vs. no) | 1.97 | (0.89, 4.37) | 0.10 | Hemoglobin level at trial entry (per g/dL) | 0.53 | (0.33, 0.87) | 0.01 | ||
| Living with a smoker at trial entry (yes vs. no) | 6.25 | (1.41, 27.80) | 0.02 | ||||||
| Model goodness of fit: Hosmer-Lemeshow χ2 (df=3)= 3.33, P=0.53, indicating good fit. | Ever diagnosed with cardiovascular disease as of trial entry (yes vs. no) | 0.26 | (0.07, 0.94) | 0.04 | |||||
| Current tobacco smoker as of trial entry (yes vs. no) | 0.36 | (0.11, 1.15) | 0.08 | ||||||
| Ever diagnosed with osteoarthritis, sciatica, chronic back problems as of trial entry (yes vs. no) | 0.41 | (0.15, 1.14) | 0.09 | ||||||
| Exacerbation in 3 mos prior to trial entry (yes vs. no) | 3.61 | (0.82, 15.76) | 0.09 | ||||||
| Model goodness of fit: Hosmer-Lemeshow χ2 (df=3)= 1.24, P=0.74, indicating good fit. | |||||||||
| Long term (month 13 to end of follow-up; N=187) | Long term (month 13 to end of follow-up; N=133) | ||||||||
| Age at trial entry (per year) | 1.08 | (1.01, 1.16) | 0.02 | Age at trial entry (per year) | 0.96 | (0.89, 1.04) | 0.34 | ||
| White race (yes vs. no) | 1.46 | (0.46, 4.58) | 0.52 | White race (vs. other) | 5.35 | (0.95, 30.06) | 0.06 | ||
| Female (vs. male) | 2.04 | (0.71, 5.84) | 0.19 | Female (vs. male) | 2.11 | (0.44, 10.23) | 0.35 | ||
| High adherence in months 9–12 (yes vs. no) | 12.49 | (4.90, 31.79) | <0.001 | High adherence in months 9–12 (yes vs. no) | 38.08 | (6.96, 208.20) | <0.001 | ||
| High adherence in the first 30 days (yes vs. no) | 2.95 | (1.17, 7.43) | 0.02 | High adherence in the 1st 30 days (yes vs. no) |
3.33 | (0.91, 12.22) | 0.07 | ||
| 6-minute walk distance at trial entry (per 10 feet) | 0.99 | (0.98, 1.01) | 0.20 | 6-minute walk distance at trial entry (per 10 feet) | 0.98 | (0.96, 1.00) | 0.05 | ||
| SGRQ total score at trial entry (per unit) | 1.03 | (1.01, 1.06) | 0.01 | Hospitalized in year prior to trial entry (yes vs. no) | 20.22 | (1.54, 266.25) | 0.02 | ||
| Hemoglobin level at trial entry (per g/dL) | 1.37 | (1.00, 1.88) | 0.05 | Ever diagnosed with osteoarthritis, sciatica, chronic back problems as of trial entry (yes vs. no) | 0.23 | (0.06, 0.89) | 0.03 | ||
| Self-reported rating of readiness to use oxygen at randomization (per unit) | 1.18 | (0.98, 1.41) | 0.08 | Epworth Sleepiness Scale score at trial entry (per unit) | 1.18 | (0.98, 1.42) | 0.08 | ||
| Post BD FEV1 at trial entry (per liter) | 2.32 | (0.87, 6.20) | 0.09 | ||||||
| Model goodness of fit: Hosmer-Lemeshow χ2 (df=3)= 5.25, P=0.15, indicating good fit. | Model goodness of fit: Hosmer-Lemeshow χ2 (df=3)= 3.40, P=0.33, indicating good fit. | ||||||||
For each outcome, the model shown is that with the lowest Akaike information criterion (16, 17). Odds ratios, 95% confidence intervals, and P values (Wald chi-squared tests) were determined from multiple logistic regression of high adherence in the indicated time frame (yes vs. no) on the candidate set of characteristics. For each model, inclusion of age, race and sex was forced; the other characteristics were identified by best variable subsets selection (15) using the Akaike information criterion and were selected from the candidate set of characteristics for the outcome (20 for short-term adherence; 22 for medium-term adherence; and 23 for long-term adherence). See e-Appendix for additional details.
LTOT=long-term oxygen treatment; MMRC = modified Medical Research Council; BD=bronchodilator; FEV1 = forced expiratory volume in 1 second; SGRQ = St. George’s Respiratory Questionnaire.