Table 3. Logistic regression to identify predictors of concurrent chemoradiotherapy versus single-modality therapy among patients with unresected, stage III NSCLC patients.
| Patient Characteristics | Logistic regression selected variables1 | |
|---|---|---|
| Odds Ratio (95% CI) | P-value | |
| Age | 0.93 (0.92, 0.94) | <0.001* |
| Female | 0.85 (0.74, 0.99) | 0.034* |
| Race | ||
| White | Ref | |
| Asian | 1.13 (0.74, 1.73) | 0.568 |
| Black | 0.71 (0.55, 0.91) | 0.006* |
| Other | 0.99 (0.61, 1.59) | 0.963 |
| Specialists visited2 | ||
| One type | Ref | |
| Two types | 1.16 (0.98, 1.36) | 0.081 |
| Three types | 1.87 (1.51, 2.32) | <0.001* |
| Region | ||
| West | Ref | |
| Midwest | 1.59 (1.25, 2.04) | <0.001* |
| Northeast | 1.25 (1.01, 1.54) | 0.038* |
| South | 1.06 (0.86, 1.29) | 0.589 |
| Setting of residence | ||
| Big metro | Ref | |
| Metro | 1.20 (1.01, 1.42) | 0.035* |
| Urban | 1.15 (0.84, 1.58) | 0.370 |
| Less Urban | 1.18 (0.90, 1.54) | 0.244 |
| Rural | 1.16 (0.74, 1.84) | 0.515 |
| Primary tumor site | ||
| Upper lobe | Ref | |
| Lower lobe | 0.72 (0.61, 0.86) | <0.001* |
| Lung, unspecified | 0.63 (0.48, 0.83) | <0.001* |
| Main bronchus | 1.30 (0.94, 1.80) | 0.106 |
| Mid-lobe | 1.42 (0.96, 2.11) | 0.078 |
| Overlap | 0.66 (0.31, 1.42) | 0.285 |
| Grade | ||
| Well differentiated | Ref | |
| Moderately differentiated | 1.31 (0.80, 2.14) | 0.277 |
| Poorly differentiated | 1.63 (1.01, 2.62) | 0.045* |
| Undifferentiated | 2.41 (1.01, 5.76) | 0.048* |
| Unknown | 1.27 (0.79, 2.03) | 0.324 |
| Histology | ||
| Adenocarcinoma | Ref | |
| Squamous cell | 1.28 (1.08, 1.52) | 0.005* |
| General NSCLC | 1.16 (0.92, 1.48) | 0.215 |
| Other | 1.10 (0.83, 1.45) | 0.519 |
| Stage | ||
| 3A | Ref | |
| 3B | 0.83 (0.72, 0.97) | 0.020* |
| Index year of first therapeutic regimen | ||
| 2009 | Ref | |
| 2010 | 1.13 (0.90, 1.42) | 0.285 |
| 2011 | 1.32 (1.05, 1.65) | 0.019* |
| 2012 | 1.54 (1.22, 1.95) | <0.001* |
| 2013 | 1.43 (1.13, 1.82) | 0.003* |
| 2014 | 0.79 (0.40, 1.56) | 0.498 |
| Metastatic site codes3 | ||
| Lip, oral cavity, and pharynx | 2.02 (0.78, 5.23) | 0.145 |
| Lymphatic and hematopoietic tissue | 1.91 (1.02, 3.56) | 0.043* |
| Genitourinary organs | 1.71 (1.39, 2.12) | <0.001* |
| Bone and bone marrow | 0.51 (0.34, 0.76) | <0.001* |
| Adrenal gland | 0.42 (0.14, 1.25) | 0.119 |
| Charlson Comorbidity Index | 0.90 (0.86, 0.94) | <0.001* |
| Comorbidities present during baseline period | ||
| Dyslipidemia | 1.37 (1.17, 1.61) | <0.001* |
| COPD | 1.04 (0.87, 1.25) | 0.654 |
| Ischemic heart disease | 1.03 (0.88, 1.21) | 0.693 |
| Chronic kidney disease | 0.87 (0.68, 1.10) | 0.242 |
| Heart failure | 0.78 (0.63, 0.98) | 0.032* |
| Proportion of patients with ≥1 tumor marker tests | 0.86 (0.63, 1.17) | 0.345 |
| Predicted performance status | ||
| Poor | Ref | |
| Good | 1.72 (1.46, 2.02) | <0.001* |
Abbreviations: NSCLC: non-small cell lung cancer; CRT: Chemoradiotherapy; CI: Confidence interval
* Significant at the 5% level
[1] A stepwise backwards elimination technique was used to identify significant baseline characteristics with a p-value cutoff of 0.20 for variable removal. Odds ratios were estimated from a logistic regression model adjusting for the following baseline covariates: race, gender, age, region, setting of residence, primary tumor site, grade, laterality, histology, index year of first therapeutic regimen, number of unique metastatic sites, all cancer metastatic sites at baseline, all comorbidities present at baseline, CCI, tumor marker tests, and predicted performance status.
[2] Specialists included were medical oncologists, radiation oncologists, and surgery specialists.
[3] Identified based on the following ICD-9 codes: 140–149, 170, 179–189, 194.0, 196, 198.5, 198.6, 198.7, 198.82, 200–208.