Table 2.
Adjusted Odds of Receiving Treatment with an Oral Anticoagulant for Resident, Nursing Home, and County Characteristics Estimated from Multilevel Logistic Models with Random County and State Intercepts
| Adjusted Odds Ratio* | |
|---|---|
| Year (reference: 2014) | |
| 2015 | 1.04 (1.00-1.08) |
| 2016 | 1.18 (1.14-1.22) |
| Demographics | |
| Age (1-year increase from mean 83.7 years) | 0.97 (0.97-0.97) |
| Women | 0.98 (0.94-1.02) |
| Married | 1.09 (1.05-1.13) |
| Medicaid eligible | 1.00 (0.96-1.04) |
| Race/ethnicity (reference: Non-Hispanic White) | |
| Non-Hispanic Black/African American | 1.00 (0.94-1.06) |
| Hispanic | 0.90 (0.80-1.02) |
| Asian/Pacific Islander | 0.72 (0.63-0.82) |
| Other/Unknown | 0.92 (0.81-1.05) |
| Days since first observed nursing home admission (Q4 vs. Q1) | 1.06 (1.01-1.10) |
| Hospital admissions in prior year (reference: 0) | |
| 1 | 0.98 (0.90-1.07) |
| 2 | 0.98 (0.90-1.07) |
| 3 | 0.97 (0.88-1.06) |
| 4+ | 0.98 (0.89-1.08) |
| Ischemic stroke | 1.49 (1.40-1.59) |
| Transient ischemic attack | 1.38 (1.22-1.56) |
| Extracranial bleeding | 0.63 (0.59-0.66) |
| Intracranial hemorrhage | 0.26 (0.22-0.31) |
| Venous thromboembolism | 2.88 (2.63-3.16) |
| Acute myocardial infarction | 0.87 (0.80-0.93) |
| At least 1 inpatient surgical procedure in past 6 months | 0.96 (0.93-0.99) |
| Unique medications (Q4 vs. Q1) | 1.38 (1.32-1.45) |
| Select Medications, † | |
| Nonsteroidal anti-inflammatory drug | 0.87 (0.84-0.91) |
| Antiplatelet | 0.32 (0.30-0.33) |
| Statin | 1.41 (1.37-1.45) |
| Selective serotonin reuptake inhinitor | 1.01 (0.98-1.04) |
| Angiotensin converting enzyme inhibitor/angiotensin receptor blocker | 1.10 (1.07-1.14) |
| Select comorbidities,‡ | |
| Diabetes mellitus | 1.09 (1.06-1.13) |
| Heart failure | 1.29 (1.25-1.33) |
| Hypertension | 1.09 (1.05-1.14) |
| Coronary artery disease | 0.92 (0.89-0.95) |
| Peripheral vascular disease | 1.23 (1.18-1.29) |
| Anemia | 0.79 (0.76-0.81) |
| Fall history | |
| Fall with fracture in six months before last admission | 0.92 (0.80-1.06) |
| Fall since admission | 0.86 (0.83-0.89) |
| Hip fracture | 0.99 (0.91-1.09) |
| Stroke | 1.17 (1.12-1.23) |
| Aphasia | 1.14 (1.06-1.24) |
| Hemiplegia | 1.48 (1.40-1.56) |
| Malnutrition | 0.80 (0.74-0.87) |
| Liver cirrhosis | 0.51 (0.42-0.63) |
| Cancer | 0.83 (0.78-0.88) |
| Coagulopathy | 1.09 (1.01-1.18) |
| Renal impairment (reference: none) | |
| Chronic renal insufficiency | 0.96 (0.93-1.00) |
| End-stage renal disease | 0.88 (0.85-0.92) |
| Dialysis | 0.75 (0.69-0.82) |
| CHA2DS2-Vasc Risk Score,‡ (reference: 2–3) | |
| 4 | 1.28 (1.21-1.34) |
| 5 | 1.51 (1.44-1.59) |
| 6 | 1.73 (1.63-1.83) |
| 7+ | 2.08 (1.96-2.21) |
| ATRIA Bleeding Risk Score,‡ (reference: low 0-3) | |
| Intermediate (4) | 0.69 (0.65-0.73) |
| High (5-10) | 0.75 (0.72-0.77) |
| Level of cognitive impairment (reference: none) | |
| Mildly impaired | 0.87 (0.83-0.90) |
| Moderately impaired | 0.69 (0.66-0.71) |
| Severely impaired | 0.49 (0.45-0.52) |
| Activites of daily living score (reference: 0-4) | |
| 5-8 | 0.98 (0.94-1.03) |
| 9-12 | 0.96 (0.92-1.01) |
| 13-16 | 0.82 (0.77-0.88) |
| Nursing Home Characteristics | |
| Number of beds (Q4 vs. Q1) | 0.98 (0.93-1.03) |
| Occupancy(Q4 vs. Q1) | 1.01 (0.96-1.06) |
| Ownership (reference: government) | |
| For profit, individual/partner entity | 1.05 (0.97-1.13) |
| For profit corporation | 1.06 (0.99-1.14) |
| Non-profit church or other non-corporation | 1.12 (1.02-1.22) |
| Non-profit corporation | 1.14 (1.06-1.23) |
| Nursing home compare overall rating,§ (reference: 1) | |
| 2 | 0.99 (0.94-1.04) |
| 3 | 1.03 (0.98-1.09) |
| 4 | 1.05 (0.99-1.10) |
| 5 | 1.07 (1.02-1.13) |
| Clinical lab available | 1.02 (0.98-1.06) |
| Medical director | 1.01 (0.95-1.06) |
| Physician and extender minutes/resident/day (Q4 vs. Q1) | 0.96 (0.91-1.02) |
| Proportion of minutes from physician extenders (Q4 vs. Q1) | 1.00 (0.95-1.05) |
| Nursing minutes/resident/day (Q4 vs. Q1) | 1.00 (0.95-1.05) |
| Proportion of minutes from registered nurses (Quartile 4 vs. 1) | 1.06 (1.01-1.11) |
| Pharmacist minutes/resident/day (Q4 vs. Q1) | 1.01 (0.97-1.05) |
| Hospice beds | 0.92 (0.78-1.07) |
| Special rehabilitation services | 0.92 (0.84-1.01) |
| County characteristics | |
| Area sociodemographics | |
| Proportion of adults ≥65 eligible for Medicaid (Q4 vs. Q1) | 1.11 (1.01-1.21) |
| Proportion of adults ≥25 years of age without high school diploma (Q4 vs. Q1) | 1.09 (1.00 -1.18) |
| Non-white race/ethnicity, (Q4 vs. Q1) | 0.87 (0.79-0.96) |
| Adults ≥65 years of age in deep poverty (Q4 vs. Q1) | 0.97 (0.91-1.04) |
| Single parent households per 10,000 persons (Q4 vs. Q1) | 0.94 (0.86-1.01) |
| Urban-rural continuum (reference: metro area) | |
| Urban area metro adjacent | 1.02 (0.96-1.09) |
| Urban not adjacent to metro | 1.05 (0.97-1.13) |
| Rural | 1.18 (1.01-1.39) |
| Area healthcare resources | |
| Hospitals | |
| Hospitals per 100 square miles (Q4 vs. Q1) | 1.08 (0.98-1.19) |
| At least 1 hospital with geriatric services, Hospitals affiliated with a medical school | 0.98 (0.93-1.03) |
| One medical school affiliated hospital in the county | 0.97 (0.92-1.03) |
| ≥2 medical school affiliated hospitals in the county | 0.98 (0.91-1.05) |
| Providers | |
| Cardiologists per 10,000 persons, median (Q4 vs. Q1) | 0.99 (0.91-1.07) |
| At least 1 hospice provider | 0.93 (0.88-0.98) |
| Health | |
| Cerebrovascular deaths per 10,000 persons (Q4 vs. Q1) | 0.97 (0.90-1.04) |
*
Adjusted odds ratios were estimated from a logistic model with random intercepts for county and state and fixed effects for resident, nursing home, and county characteristics
†
Individual medication estimates derived from a model omitting the number of unique medications
‡
CHA2DS2-Vasc Risk Score and ATRIA Bleeding Risk Score adjusted estimates derived from a separate model omitting the variables included in the scores. Adjusted estimates for comorbidities included in the scores were obtained from models omitting the CHA2DS2-Vasc Risk Score and ATRIA Bleeding Risk Score.
§
Nursing home quality estimates derived from a separate model omitting professional staffing variables because staffing is incorporated in the calculation of the quality score