Abstract
Background and Purpose:
Older patients with pre-existing mild cognitive impairment (MCI) receive less evidence-based care after acute myocardial infarction, however, whether they receive less care after acute ischemic stroke (AIS) is unknown. We compared receipt of guideline-concordant procedures after AIS between older adults with pre-existing MCI and normal cognition.
Methods:
Prospective study of 591 adults ≥65 hospitalized for AIS between 2000 and 2014, and followed through 2015 using data from the nationally representative Health and Retirement Study, Medicare and American Hospital Association. We assessed pre-existing MCI (modified Telephone Interview for Cognitive Status score of 7–11) and normal cognition (score of 12–27). Primary outcome was a composite quality measure representing the number of 4 procedures (carotid imaging, cardiac monitoring, echocardiogram, and rehabilitation assessment) received within 30 days after AIS (ordinal scale with values of 0, 1, 2, 3–4).
Results:
Among survivors of AIS, 26.9% had pre-existing MCI (62.9% were women, with a mean [SD] age of 82.4 [7.7] years), and 73.1% had normal cognition (51.4% were women, with a mean age of 78.4 [7.2] years). Patients with pre-existing MCI, compared to cognitively normal patients, had 39% lower cumulative odds of receiving the composite quality measure (unadjusted cumulative odds ratio, OR, 0.61 [95% CI, 0.43–0.87]; P=0.006). However, this association became non-significant after adjusting for patient and hospital factors (adjusted cumulative OR, 0.83 [95% CI, 0.56–1.24]; P=0.37). Lower cumulative odds of receiving the composite quality measure were associated with older patient age (adjusted cumulative OR per 1-year older age, 0.97 [95% CI, 0.95–0.99]; P=0.01) and Southern hospitals (adjusted cumulative OR for South vs North, 0.54 [95% CI, 0.31–0.94]; P=0.03).
Conclusions:
Differences in receipt of guideline-concordant procedures after AIS exist between patients with pre-existing MCI and normal cognition. These differences were largely explained by patient and regional factors associated with receiving less AIS care.
Keywords: ischemic stroke, cognitive impairment, aging, health policy/outcomes research
Subject terms: quality and outcomes
Introduction
Mild cognitive impairment (MCI)—measurable cognitive impairment that doesn’t severely affect daily functioning—is common, affecting 1 in 5 older adults (65+).1 The number of older adults diagnosed with MCI is projected to surge because of population aging and new coverage of an assessment of cognitive impairment in the annual wellness benefit for all Medicare beneficiaries.2 MCI does not inevitably progress to dementia even after a decade.3 Many older adults with MCI live years—an average of 9 years in one community-based study4—with good quality of life5, and so face competing health risks of aging, particularly cardiovascular disease (CVD).4 Among community-dwelling older adults with and without MCI, CVD is the leading cause of death and serious morbidity.4 Little is known about the care of patients with pre-existing MCI after acute ischemic stroke (AIS) despite evidence in other diseases, including acute myocardial infarction,6 that these patients might receive less care.
We compared receipt of established procedures after AIS between patients with pre-existing MCI and cognitively normal patients in a nationally representative sample of US older adults, and determined whether differences persisted after adjusting for patient and hospital factors.
Methods
Study Population
The Health and Retirement Study (HRS) is a nationally representative longitudinal study of 37,000 US adults 51+ evaluating health and economic changes associated with aging.7 The HRS samples from all contiguous US states, with oversampling of Blacks and Hispanics. Every 2 years since 1992, HRS participants have been interviewed using standard instruments. Trained survey interviewers from the Institute for Social Research at the University of Michigan conduct the interviews. The HRS achieves a high follow-up rate, ranging 85%–91% from 1998–2016 including proxies7, and follows participants after entry into nursing homes. Utilization and dates of inpatient and outpatient medical services were available from the Centers for Medicare and Medicaid Services (CMS) for HRS participants who were enrolled in Medicare fee-for-service (at age 65 or disability) and consented to this linkage (80–90% consent).
We identified dementia-free participants 65+ hospitalized for AIS between January 1, 2000 and December 31, 2014 using valid International Classification of Diseases-Ninth (ICD-9) Revision-CM codes (433.x1, 434.xx, 436.xx as the primary discharge diagnosis)8 in the nationally representative Health and Retirement Study (HRS) with linked Medicare data. Codes are in eTable 1. We used the first AIS hospitalization during the study period. We used hospital data from the American Hospital Association. Dates of death were available for >95% of participants from the National Death Index and Medicare. The University of Michigan Institutional Review Board approved the research protocol. All human participants gave written informed consent.
Dependent Variables
The primary outcome was an ordinal composite quality measure representing the number of 4 procedures (carotid imaging [ultrasound, or magnetic resonance or computed tomography angiography], cardiac monitoring [outpatient cardiac event monitoring or implantable loop recorder], echocardiogram [transthoracic or transesophageal], and rehabilitation assessment [physical therapy evaluation, occupational therapy evaluation, and rehabilitation considered]) received within 30 days after AIS and measured using valid ICD-9 and Current Procedural Terminology codes (eTable 1). We selected common procedures measurable in administrative data.
Secondary outcomes included the 4 individual procedures in the composite measure, carotid revascularization (carotid stenting or carotid endarterectomy) ≤90 days, a 4-level composite brain imaging measure (none, head CT only, brain MRI only, and dual brain imaging with head CT and brain MRI), and brain MRI separately because it is not universally recommended in all AIS patients.
Main Independent Variable
Trained HRS interviewers administered cognitive function tests biennially in-person or by telephone using the Modified Telephone Interview for Cognitive Status (TICS-m).9 Separate assessments were used to evaluate the cognitive function of respondents represented by a proxy.10 At each interview, the HRS participant was classified as having normal cognition (TICS-m score of 12–27), MCI (TICS-m score of 7–11), or dementia (TICS-m score <7) using validated cut-points on TICS-m and proxy assessments 10, 11 with cut-points based on in-depth, in-home, neuropsychological and clinical assessments as well as expert clinician adjudication from the Aging, Demographics, and Memory Study, an HRS dementia sub-study.12
At each interview, the HRS participant was classified as having normal cognition, MCI, or dementia using validated cut-points on the Modified Telephone Interview for Cognitive Status (TICS-m)9 or proxy assessments.10, 11
Covariates
Covariates included patient and hospital factors. Patient factors included age, sex, race/ethnicity (non-Hispanic white vs Hispanic), education (<12 years, 12 years, 13–15 years, and ≥16 years), net wealth (quartiles) (≤$51,200, $51,201–$147,000, $147,001–$356,000, and >$356,000), income (quartiles) (≤$15,580, $15,581–$26,760, $26,761–$45,980, and >$45,980), and Charlson comorbidity index score. Depressive symptoms were measured by the 8-item Center for Epidemiological Studies Depression Scale (CES-D) and categorized (0, 1–4, and 5–8). Functional limitations were measured by summing the number of difficulties with 6 activities of daily living (walking, dressing, bathing, eating, getting in/out of bed, using toilet), and 5 instrumental activities of daily living (managing money, taking medication, preparing hot meals, using phones, and shopping for groceries) (range, 0–11). Social support/structure was measured by marital status/living arrangement (married/partner, unmarried/living with other, and unmarried/living alone), geographic proximity of adult children (no children or missing, co-residence, within 10 miles, and greater than 10 miles), and having an adult daughter based on prior research.13, 14 Hospital factors were medical school affiliate or teaching hospital, region (Northeast, Midwest, South, and West), bed size (<200 beds, 200–399 beds, and ≥400 beds), and authority type (government nonfederal [public], non-profit, and for profit).
Statistical Analysis
We estimated associations between pre-existing MCI and the outcomes using ordinal logistic regression for the 4-level composite quality measure, logistic regression for binary secondary outcomes, and multinomial logistic regression for the 4-level composite brain imaging outcome, before and after adjusting for patient and hospital factors. We tested and did not find evidence that the proportional odds assumption was not met. We used single imputation for missing depressive symptoms (n=31) using all available prior depressive symptom scores. All statistical analyses were performed with STATA software, version 15.0 (StataCorp, College Station, TX).
Results
Figure 1 shows the derivation of the cohort. The study sample included 591 participants hospitalized for AIS, 159 (26.9%) of whom were classified as having pre-existing MCI. Table 1 presents patient and hospital characteristics by cognitive status. Among the patients with pre-existing MCI, the median time between the cognitive assessment of MCI and the AIS hospitalization was 427 days (interquartile range, 222–655 days). Patients with pre-existing MCI had higher one-year mortality than cognitively normal patients (31.5% versus 11.3%; P<0.001). eTable 2 compares characteristics between included and excluded patients.
Figure 1:
Flow Diagram of Derivation of Participant Cohort
*Not enrolled in Medicare Parts A and B fee-for service 12-months before and after enrolment provided the enrolees were alive for 12 months after AIS hospitalization
Abbreviation: AIS is acute ischemic stroke
We excluded those who died during the hospitalization, those who died within 7 days of hospital discharge, and those with extreme lengths of stay for index stroke hospitalization (>30 days suggesting extreme disease severity) because they are not eligible for the procedures.
Table 1:
Participant and Hospital Characteristics by Cognitive Status
| Participants with Normal Cognition (n=432) | Participants with Pre-existing MCI (n=159) | P-value | ||
|---|---|---|---|---|
| Participant characteristics | ||||
| Age at AIS, mean (SD), years | 78.4 (7.2) | 82.4 (7.7) | <0.001 | |
| Women | 222 (51.4) | 100 (62.9) | 0.01 | |
| Non-Hispanic White | 369 (85.4) | 125 (78.6) | 0.05 | |
| Marital status/living arrangement N(%) | <0.001 | |||
| Married/partner | 245 (56.7) | 58 (36.5) | ||
| Unmarried/living with other | 59 (13.7) | 32 (20.1) | ||
| Unmarried/living alone | 128 (29.6) | 69 (43.4) | ||
| Education | <0.001 | |||
| <12 years | 92 (21.3) | 64 (40.3) | ||
| 12 years | 154 (35.7) | 61 (38.4) | ||
| 13–15 years | 100 (23.2) | 24 (15.1) | ||
| ≥16 years | 86 (19.9) | 10 (6.3) | ||
| Net wealth (quartiles) | 0.002 | |||
| ≤$51,200 | 98 (22.7) | 58 (36.5) | ||
| $51,201–$147,000 | 94 (21.8) | 36 (22.6) | ||
| $147,001–$356,000 | 109 (25.2) | 36 (22.6) | ||
| >$356,000 | 131 (30.3) | 29 (18.2) | ||
| Income (quartiles) | <0.001 | |||
| ≤$15,580 | 97 (22.5) | 60 (37.7) | ||
| $15,581–$26,760 | 98 (22.7) | 43 (27.0) | ||
| $26,761–$45,980 | 117 (27.1) | 29 (18.2) | ||
| >$45,980 | 120 (27.8) | 27 (17.0) | ||
| Charlson comorbidity index score, mean (SD), points | 2.6 (2.3) | 3.0 (2.6) | 0.06 | |
| CES-D depressive symptoms | 0.006 | |||
| 0 | 178 (41.2) | 44 (27.7) | ||
| 1–4 | 206 (47.7) | 88 (55.4) | ||
| 5–8 | 48 (11.1) | 27 (17.0) | ||
| Functional limitations, mean (SD) | 0.8 (1.5) | 2.0 (2.4) | <0.001 | |
| Geographic proximity to adult children | 0.14 | |||
| No children or missing | 30 (6.9) | 10 (6.3) | ||
| Co-residence | 72 (16.7) | 33 (20.8) | ||
| Within 10 miles | 184 (42.6) | 77 (48.4) | ||
| Greater than 10 miles | 146 (33.8) | 39 (24.5) | ||
| Have an adult daughter | 348 (80.6) | 123 (77.4) | 0.39 | |
| Hospital characteristics | ||||
| Medical school affiliate or teaching hospital | 204 (47.2) | 62 (39.0) | 0.08 | |
| Region | 0.36 | |||
| Northeast | 57 (13.2) | 17 (10.7) | ||
| Midwest | 214 (49.5) | 70 (44.0) | ||
| South | 112 (25.9) | 50 (31.5) | ||
| West | 49 (11.3) | 22 (13.8) | ||
| Bed size | 0.09 | |||
| <200 beds | 120 (27.8) | 56 (35.2) | ||
| 200–399 beds | 157 (36.3) | 60 (37.7) | ||
| ≥400 beds | 155 (35.9) | 43 (27.0) | ||
| Authority type | 0.70 | |||
| Government nonfederal (public) | 40 (9.3) | 16 (10.1) | ||
| Non-profit | 319 (73.8) | 112 (70.4) | ||
| For profit | 73 (16.9) | 31 (19.5) | ||
Abbreviations: MCI is mild cognitive impairment. AIS is acute ischemic stroke. CES-D is Center for Epidemiological Studies Depression Scale. SD is standard deviation. CT is computed tomography. MRI is magnetic resonance imaging.
Functional limitations were measured by summing the number of difficulties with 6 activities of daily living (walking, dressing, bathing, eating, getting in/out of bed, using toilet), and 5 instrumental activities of daily living (managing money, taking medication, preparing hot meals, using phones, and shopping for groceries) (range, 0–11).
Depressive symptoms were measured by the 8-item Center for Epidemiological Studies Depression Scale.
Social support/structure was measured by marital status/living arrangement, having a daughter, and geographic proximity of adult children.
P-values were calculated using chi square test for categorical variables and t-test for continuous variables.
Patient characteristics were available from the Health and Retirement Study.
Hospital characteristics were available from the American Hospital Association.
Patients with pre-existing MCI were less likely to receive the composite quality measure than cognitively normal patients within 30 days after AIS (Table 2). Few patients with pre-existing MCI and normal cognition received 3–4 of the 4 procedures in the composite quality score. In unadjusted analysis, patients with pre-existing MCI had 39% lower cumulative odds of receiving the composite quality measure within 30 days after AIS (Table 2). However, pre-existing MCI was not independently associated with receipt of the composite quality measure after adjusting for patient and hospital factors (Table 2).
Table 2:
Receipt of composite quality measure within 30 days between participants with pre-existing MCI and cognitively normal patients
| Number of 4 procedures within composite quality measure received | Participants with Pre-existing Normal Cognition (n=432) | Participants with Pre-existing MCI (n=159) | P-value | Unadjusted cumulative odds ratios (95% CI) for MCI vs normal cognition | Adjusted cumulative odds ratios (95% CI) for MCI vs normal cognition |
|---|---|---|---|---|---|
| 0 | 40 (9.3) | 19 (12.0) | 0.03 | 0.61 (0.43–0.87) P=0.006 |
0.83 (0.56–1.24) P=0.37 |
| 1 | 88 (20.4) | 48 (30.2) | |||
| 2 | 257 (59.5) | 80 (50.3) | |||
| 3–4 | 47 (10.9) | 12 (7.6) |
Abbreviations: MCI, mild cognitive impairment.
The primary outcome was a composite quality measure representing the number of 4 procedures (carotid imaging, cardiac monitoring, echocardiogram, and rehabilitation assessment) received within 30 days after AIS (ordinal scale with values of 0, 1, 2, 3–4).
We combined the 2 categories corresponding to receiving 3 and 4 of the 4 procedures because only one individual received all 4 procedures resulting in a 4-level composite quality measure (values of 0, 1, 2, 3–4). Ordinal logistic regression models estimated the cumulative odds of receiving the composite quality measure (primary outcome) before and after adjusting for patient factors (age, sex, race/ethnicity, education, net wealth, income, Charlson comorbidity index score, depressive symptoms, functional limitations in basic and instrumental activities of daily living, marital status/living arrangement, geographic proximity to adult children, and having an adult daughter) and hospital factors (medical school affiliate or teaching hospital, region, bed size, and authority type).
Results of the secondary outcomes were consistent though no contrasts between cognitive status groups were statistically significant (Table 3).
Table 3:
Receipt of Individual Procedures (Secondary Outcomes) within 30 Days After Acute Ischemic Stroke Between Participants with Pre-existing Mild Cognitive Impairment and Cognitively Normal Patients
| Receipt of individual procedures (secondary outcomes) | Participants with Normal Cognition (n=432) | Participants with Pre-existing MCI (n=159) | P-value* | Unadjusted odds ratios (95% CI) for MCI vs normal cognition | Adjusted odds ratios (95% CI) for MCI vs normal cognition |
|---|---|---|---|---|---|
| Carotid imaging | 342 (79.2) | 118 (74.2) | 0.20 | 0.76 (0.50–1.16) P=0.20 |
0.93 (0.58–1.51) P=0.78 |
| Cardiac monitoring | 24 (5.6) | 3 (1.9) | 0.06 | 0.33 (0.10–1.10) P=0.07 |
NA |
| Echocardiogram | 332 (76.9) | 113 (71.1) | 0.15 | 0.74 (0.49–1.11) P=0.15 |
0.93 (0.58–1.50) P=0.77 |
| Brain MRI | 247 (57.2) | 83 (52.2) | 0.28 | 0.82 (0.57–1.18) P=0.28 |
1.13 (0.74–1.74) P=0.56 |
| Rehabilitation assessment | 46 (10.7) | 10 (6.3) | 0.11 | 0.56 (0.28–1.14) P=0.11 |
NA |
| Carotid revascularization** | 22 (5.1) | 3 (1.9) | 0.09 | 0.36 (0.11–1.21) P=0.10 |
NA |
| Composite brain imaging measure | |||||
| None | 45 (10.4) | 14 (8.8) | 0.38 | Referent | Referent |
| Head CT only | 140 (32.4) | 62 (39.0) | 1.42 (0.73–2.78) P=0.30 |
1.13 (0.52–0.47) P=0.75 |
|
| Brain MRI only | 53 (12.3) | 14 (8.8) | 0.85 (0.37–1.97) P=0.70 |
0.77 (0.29–2.06) P=0.61 |
|
| Dual brain imaging with CT and MRI | 194 (44.9) | 69 (43.4) | 1.14 (0.59–2.21) P=0.69 |
1.39 (0.64–3.00) P=0.40 |
Abbreviations: MCI, mild cognitive impairment. CT, computed tomography. MRI, magnetic resonance imaging. NA, not applicable.
P-values were calculated using chi square test for categorical variables.
Carotid revascularization measured within 90 days after acute ischemic stroke. Carotid imaging included ultrasound, magnetic resonance angiography, or computerized tomography angiography. Cardiac monitoring included cardiac event monitor or implantable loop recorder. Echocardiogram include transthoracic and transesophageal echocardiogram. Logistic regression models estimated the odds of receiving the individual procedures (secondary outcomes) before and after adjusting for patient and hospital factors.
Adjusted models included patient factors (age, sex, race/ethnicity, education, net wealth, income, Charlson comorbidity index score, depressive symptoms, functional limitations in basic and instrumental activities of daily living, marital status/living arrangement, geographic proximity to adult children, and having an adult daughter) and hospital factors (medical school affiliate or teaching hospital, region, bed size, and authority type).
We do not present adjusted results for these individual procedures because numbers are small.
Factors associated with lower cumulative odds of receiving the composite quality measure and commonly performed procedures were older age, Southern hospital region, and smaller hospital size (<200 beds vs 400+ beds) (Table 4). We do not present adjusted results for cardiac monitoring, rehabilitation assessment, and carotid revascularization because numbers are small.
Table 4:
Predictors of Receipt of Composite Quality Measure and Individual Procedures within 30 Days after Acute Ischemic Stroke (Adjusted Odds Ratios and 95% Confidence Intervals) (n=591)
| Primary Outcome | Secondary Outcomes | |||
|---|---|---|---|---|
| Composite Quality Measure | Carotid imaging | Echocardiogram | Brain MRI | |
| Predictor | Cumulative Odds Ratios (95% CI) | Odds Ratios (95% CI) | Odds Ratios(95% CI) | Odds Ratios (95% CI) |
| Participant characteristics | ||||
| Pre-existing MCI vs normal cognition | 0.83 (0.56–1.24) P=0.37 |
0.93 (0.58–1.51) P=0.78 |
0.93 (0.58–1.50) P=0.77 |
1.13 (0.74–1.74) P=0.56 |
| Age per 1-year increase |
0.97 (0.95–0.99) P=0.01 |
0.97 (0.94–1.00) P=0.03 |
0.97 (0.94–1.00) P=0.03 |
0.94 (0.92–0.97) P<0.001 |
| Women vs men | 0.97 (0.67–1.39) P=0.86 |
1.12 (0.70–1.78) P=0.63 |
0.95 (0.60–1.49) P=0.82 |
0.94 (0.63–1.40) P=0.77 |
| White vs non-White | 0.98 (0.61–1.57) P=0.92 |
1.13 (0.63–2.03) P=0.69 |
0.72 (0.39–1.34) P=0.30 |
1.23 (0.73–2.06) P=0.43 |
| Unmarried/living with other vs married | 0.94 (0.52–1.68) P=0.83 |
1.41 (0.65–3.05) P=0.38 |
1.12 (0.54–2.35) P=0.76 |
1.58 (0.82–3.05) P=0.17 |
| Unmarried/living alone vs married | 1.16 (0.75–1.80) P=0.51 |
0.93 (0.54–1.60) P=0.78 |
1.40 (0.82–2.38) P=0.22 |
1.48 (0.92–2.37) P=0.10 |
| High school graduate vs <high school | 1.29 (0.83–2.00) P=0.26 |
0.99 (0.58–1.72) P=0.98 |
1.25 (0.73–2.14) P=0.42 |
0.93 (0.58–1.50) P=0.77 |
| Some college vs <high school | 1.06 (0.64–1.76) P=0.82 |
0.86 (0.46–1.60) P=0.63 |
0.79 (0.43–1.45) P=0.46 |
0.91 (0.53–1.57) P=0.74 |
| College+ vs <high school | 1.00 (0.56–1.78) P=0.99 |
1.01 (0.49–2.11) P=0.97 |
0.91 (0.45–1.84) P=0.79 |
0.84 (0.45–1.56) P=0.58 |
| Wealth $51,201–$147,000 vs <$51,201 | 0.86 (0.53–1.41) P=0.55 |
1.23 (0.66–2.29) P=0.51 |
0.89 (0.49–1.63) P=0.71 |
1.10 (0.65–1.87) P=0.71 |
| Wealth $147,001–$356,000 vs <$51,201 | 0.89 (0.53–1.49) P=0.66 |
1.18 (0.62–2.24) P=0.62 |
0.81 (0.44–1.51) P=0.51 |
1.88 (1.08–3.26) P=0.03 |
| Wealth >$356,001 vs <$51,201 | 0.93 (0.54–1.60) P=0.79 |
0.79 (0.40–1.55) P=0.49 |
1.09 (0.56–2.16) P=0.79 |
1.30 (0.72–2.32) P=0.39 |
| Income $15,581–$26,760 vs <$15,581 | 0.96 (0.59–1.55) P=0.86 |
0.93 (0.52–1.66) P=0.80 |
0.86 (0.48–1.54) P=0.61 |
1.04 (0.62–1.75) P=0.88 |
| Income $26,761–$45,980 vs <$15,581 | 1.00 (0.59–1.69) P=0.99 |
1.16 (0.60–2.24) P=0.66 |
0.85 (0.45–1.60) P=0.62 |
0.76 (0.43–1.33) P=0.34 |
| Income >$45,980 vs <$15,581 | 1.46 (0.82–2.60) P=0.19 |
1.71 (0.82–3.60) P=0.16 |
1.32 (0.65–2.69) P=0.45 |
1.08 (0.58–2.00) P=0.81 |
| Charlson comorbidity index score per one point increase | 0.96 (0.89–1.02) P=0.20 |
1.00 (0.91–1.08) P=0.91 |
0.92 (0.85–1.00) P=0.05 |
0.95 (0.88–1.02) P=0.16 |
| 1–4 depressive symptoms vs none | 0.82 (0.56–1.18) P=0.28 |
0.81 (0.50–1.30) P=0.38 |
0.87 (0.55–1.37) P=0.54 |
0.65 (0.43–0.97) P=0.03 |
| 4–8 depressive symptoms vs none | 0.66 (0.37–1.17) P=0.16 |
0.58 (0.29–1.17) P=0.13 |
0.86 (0.42–1.75) P=0.67 |
0.88 (0.47–1.67) P=0.70 |
| Number of functional limitations per one unit increase | 0.97 (0.87–1.07) P=0.51 |
0.99 (0.88–1.11) P=0.83 |
0.98 (0.87–1.10) P=0.76 |
0.99 (0.89–1.11) P=0.91 |
| Co-reside with child vs no children or missing | 0.86 (0.36–2.05) P=0.73 |
0.62 (0.19–2.03) P=0.43 |
1.08 (0.37–3.13) P=0.89 |
0.64 (0.24–1.67) P=0.36 |
| Live within 10 miles from child vs no children or missing | 0.90 (0.42–1.95) P=0.79 |
0.93 (0.32–2.66) P=0.89 |
0.94 (0.36–2.41) P=0.89 |
0.90 (0.38–2.09) P=0.80 |
| Live >10 miles from child vs no children or missing | 0.91 (0.42–1.96) P=0.81 |
0.85 (0.30–2.43) P=0.77 |
1.18 (0.46–3.02) P=0.74 |
1.03 (0.44–2.40) P=0.94 |
| Have daughter vs not | 1.22 (0.76–1.94) P=0.42 |
0.78 (0.42–1.45) P=0.43 |
1.30 (0.73–2.32) P=0.37 |
1.17 (0.70–1.96) P=0.55 |
| Hospital characteristics | ||||
| Teaching hospital vs non-teaching hospital | 1.15 (0.79–1.69) P=0.47 |
0.70 (0.42–1.14) P=0.15 |
1.30 (0.82–2.08) P=0.27 |
1.27 (0.84–1.91) P=0.26 |
| Midwest vs North region | 0.66 (0.39–1.10) P=0.11 |
0.64 (0.32–1.29) P=0.22 |
0.50 (0.24–1.05) P=0.07 |
0.60 (0.34–1.06) P=0.08 |
| South vs North region |
0.54 (0.31–0.94) P=0.03 |
0.63 (0.30–1.33) P=0.23 |
0.28 (0.13–0.61) P=0.001 |
0.57 (0.31–1.06) P=0.08 |
| West vs North region | 0.73 (0.38–1.41) P=0.35 |
0.97 (0.39–2.40) P=0.95 |
0.54 (0.22–1.32) P=0.18 |
1.03 (0.50–2.15) P=0.93 |
| Bed number 200–399 vs <200 | 1.20 (0.79–1.81) P=0.39 |
1.61 (0.96–2.72) P=0.07 |
0.95 (0.58–1.56) P=0.85 |
1.78 (1.13–2.79) P=0.01 |
| Bed number 400+ vs <200 | 1.53 (0.95–2.47) 0.08 |
1.68 (0.92–3.05) P=0.09 |
1.66 (0.92–2.99) P=0.09 |
2.07 (1.24–3.46) P=0.005 |
| Hospital authority non-profit vs government non-federal (public) | 1.21 (0.67–2.17) P=0.52 |
1.88 (0.97–3.67) P=0.06 |
0.74 (0.36–1.52) P=0.41 |
1.28 (0.69–2.36) P=0.44 |
| Hospital authority for-profit vs government non-federal (public) | 1.08 (0.55–2.11) P=0.82 |
1.91 (0.87–4.21) P=0.11 |
0.90 (0.40–2.06) P=0.81 |
1.34 (0.66–2.75) P=0.42 |
Bolded figures indicate the 95% confidence intervals do not include one.
Abbreviations: MCI, mild cognitive impairment.
The primary outcome was a composite quality measure representing the number of 4 procedures (carotid imaging, cardiac monitoring, echocardiogram, and rehabilitation assessment) received within 30 days after AIS (ordinal scale with values of 0, 1, 2, 3–4).
Ordinal logistic regression models estimated the cumulative odds of receiving the composite quality measure (primary outcome) before and after including patient factors and hospital factors.
Logistic regression models estimated the odds of receiving the individual procedures (secondary outcomes) before and after adjusting for patient and hospital factors.
Adjusted models included patient factors (age, sex, race/ethnicity, education, net wealth, income, Charlson comorbidity index score, depressive symptoms, functional limitations in basic and instrumental activities of daily living, marital status/living arrangement, geographic proximity to adult children, and having an adult daughter) and hospital factors (medical school affiliate or teaching hospital, region, bed size, and authority type).
Results were similar in analyses of procedures received within 90 days after AIS (eTables 3 and 4).
Discussion
Nearly 30% of older patients had pre-existing MCI at the time of their AIS. We found no evidence that pre-existing MCI is independently associated with receiving less care after AIS.
Prior studies have shown that patients with pre-existing dementia receive less care after AIS.15, 16 Our study extends previous work by examining receipt of care in the larger group of patients with pre-existing MCI. Although patients with pre-existing MCI, compared with cognitively normal patients, received the composite quality measure less frequently after AIS, this difference was largely explained by patients with MCI being more likely to have older age and being treated at smaller and Southern hospitals, all factors associated with receiving less care.17 Physicians might recommend similar care for AIS to patients with MCI as cognitively normal patients because studies show that the two groups have similar benefits and risks of AIS treatments after controlling for patient factors including age and stroke severity.18
We found that stroke survivors who were older received worse quality of care consistent with previous research.19 We also found that stroke survivors treated at hospitals that were smaller in size and located in the South received worse quality of care. Low-quality hospitals are more likely to be smaller size, located in the South compared with the Northeast, public or for-profit hospitals (vs private nonprofit), and care for a higher percentage of elderly minorities and the elderly poor.17 Our results suggest a scientific need to better understand why patients who are older and treated at smaller and Southern hospitals receive less AIS care to inform quality improvement interventions.
Our study has several strengths. The HRS is a nationally representative study with rigorously collected data. The HRS’s longitudinal, objective measurements of cognition7 before AIS hospitalization are a unique strength because cognitive function is not routinely collected in clinical practice and MCI is under-documented in administrative data.20 We adjusted for functional limitations and comorbidity. We used a pre-specified analysis plan.
Our study has limitations. Small numbers might have limited our ability to detect differences by cognitive status. The definition of MCI is based on a limited set of cognitive tests or proxy assessments not a full clinical evaluation, so misclassification is possible, although the approach accurately classifies 85–90% of HRS self-respondents as having normal cognition versus any cognitive impairment (MCI or dementia).10, 11 We lacked information on stroke severity or functional disability after AIS, delirium, and the appropriateness of the use of AIS treatments. Guidelines recommend echocardiogram in selected patients. Dementia could have occurred between cognitive assessment and stroke hospitalization. Although administrative data might not fully capture procedures, incomplete outcome assessment would not be expected to differ by patient MCI status.
Our findings suggest that more than one in four patients with AIS has pre-existing MCI. Older adults with MCI, their loved ones, and their providers must assess the benefits and risks of stroke treatments as well as the probabilities of competing risks (e.g., death and dementia). There is no evidence that patients with MCI experience fewer benefits and greater risks of effective AIS treatments, so it is important that they receive high-quality AIS care. Our results suggest that patient MCI is not associated with receiving less care for AIS independent of patient and hospital factors in this sample. However, our findings suggest that efforts to improve AIS care in older patients and at smaller and Southern hospitals are needed. In addition, increased use of rehabilitation and cardiac monitoring after AIS in older adults with normal cognition and MCI are warranted.
Conclusions
Differences in receipt of procedures after AIS—namely use of carotid imaging, cardiac monitoring, echocardiogram, and rehabilitation assessment—exist between older patients with pre-existing MCI and those with normal cognition. These differences were largely explained by patients with MCI being more likely to have older age and being treated at smaller and Southern hospitals, all factors associated with lower quality of care.
Supplementary Material
Acknowledgements:
Sources of Funding: This work was supported by NIH/NIA grant R01 AG051827 (Levine DA, PI). The Health and Retirement Study is funded by the NIH/NIA (U01 AG009740), and performed at the Institute for Social Research, University of Michigan, Ann Arbor. Additional support includes NINDS R01 NS102715 (DAL), event adjudicator for POINT trial (DAL), NIH/NIA grants P30 AG053760 (KML), P30 AG024824 (AG, MUK, KML), and R01 AG053972 (KML).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Prior presentations: This paper was presented as an abstract at the 2019 International Stroke Conference, February 6, 2019, Honolulu, HI.
Conflicts of Interest/Disclosures: The authors declare that they do not have a conflict of interest.
Contributor Information
Deborah A. Levine, Department of Internal Medicine and Cognitive Health Services Research Program, University of Michigan, Ann Arbor, MI; Department of Neurology and Stroke Program, University of Michigan, Ann Arbor, MI; Institute for Healthcare Policy and Innovation, University of Michigan, Ann Arbor, MI.
Andrzej Galecki, Department of Internal Medicine and Cognitive Health Services Research Program, University of Michigan, Ann Arbor, MI; Department of Biostatistics, University of Michigan, Ann Arbor, MI.
Mohammed Kabeto, Department of Internal Medicine and Cognitive Health Services Research Program, University of Michigan, Ann Arbor, MI.
Brahmajee K. Nallamothu, Department of Internal Medicine and Cognitive Health Services Research Program, University of Michigan, Ann Arbor, MI; Institute for Healthcare Policy and Innovation, University of Michigan, Ann Arbor, MI; VA Ann Arbor Healthcare System, University of Michigan, Ann Arbor, MI.
Darin B. Zahuranec, Department of Neurology and Stroke Program, University of Michigan, Ann Arbor, MI.
Lewis B. Morgenstern, Department of Neurology and Stroke Program, University of Michigan, Ann Arbor, MI; Institute for Healthcare Policy and Innovation, University of Michigan, Ann Arbor, MI; Department of Epidemiology, University of Michigan, Ann Arbor, MI.
Lynda D. Lisabeth, Department of Neurology and Stroke Program, University of Michigan, Ann Arbor, MI; Department of Epidemiology, University of Michigan, Ann Arbor, MI.
Bruno Giordani, Department of Psychiatry & Michigan Alzheimer’s Disease Center, University of Michigan, Ann Arbor, MI.
Kenneth M. Langa, Department of Internal Medicine and Cognitive Health Services Research Program, University of Michigan, Ann Arbor, MI; Institute for Healthcare Policy and Innovation, University of Michigan, Ann Arbor, MI; VA Ann Arbor Healthcare System, University of Michigan, Ann Arbor, MI; Institute for Social Research, University of Michigan, Ann Arbor, MI.
References
- 1.Plassman BL, Langa KM, Fisher GG, Heeringa SG, Weir DR, Ofstedal MB, et al. Prevalence of cognitive impairment without dementia in the united states. Ann Intern Med. 2008;148:427–434 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Medicare Annual Wellness Visit. Available at: https://www.medicare.gov/coverage/yearly-wellness-visits. Accessed May 17, 2020
- 3.Mitchell AJ, Shiri-Feshki M. Rate of progression of mild cognitive impairment to dementia--meta-analysis of 41 robust inception cohort studies. Acta psychiatr Scand. 2009;119:252–265 [DOI] [PubMed] [Google Scholar]
- 4.Sachs GA, Carter R, Holtz LR, Smith F, Stump TE, Tu W, et al. Cognitive impairment: An independent predictor of excess mortality: A cohort study. Ann Intern Med. 2011;155:300–308 [DOI] [PubMed] [Google Scholar]
- 5.Barrios H, Narciso S, Guerreiro M, Maroco J, Logsdon R, de Mendonca A. Quality of life in patients with mild cognitive impairment. Aging Ment Health. 2013;17:287–292 [DOI] [PubMed] [Google Scholar]
- 6.Levine DA, Langa KM, Galecki A, Kabeto M, Morgenstern LB, Zahuranec DB, et al. Mild cognitive impairment and receipt of treatments for acute myocardial infarction in older adults. J Gen Intern Med. 2020;35:28–35 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sonnega A, Faul JD, Ofstedal MB, Langa KM, Phillips JW, Weir DR. Cohort profile: The Health and Retirement Study (HRS). Int J Epidemiol. 2014;43:576–585 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tirschwell DL, Longstreth WT Jr. Validating administrative data in stroke research. Stroke. 2002;33:2465–2470 [DOI] [PubMed] [Google Scholar]
- 9.Welsh KA, Breitner JCS, Magruderhabib KM. Detection of dementia in the elderly using telephone screening of cognitive status. Neuropsychiatry Neuropsychol Behav Neurol. 1993;6:103–110 [Google Scholar]
- 10.Crimmins EM, Kim JK, Langa KM, Weir DR. Assessment of cognition using surveys and neuropsychological assessment: The Health and Retirement Study and the Aging, Demographics, and Memory Study. J Geront B Pyschol Sci Soc Sci. 2011;66 Suppl 1:i162–171 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Langa KM, Larson EB, Crimmins EM, Faul JD, Levine DA, Kabeto MU, et al. A comparison of the prevalence of dementia in the United States in 2000 and 2012. JAMA Intern Med. 2017;177:51–58 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Langa KM, Plassman BL, Wallace RB, Herzog AR, Heeringa SG, Ofstedal MB, et al. The Aging, Demographics, and Memory Study: Study design and methods. Neuroepidemiology. 2005;25:181–191 [DOI] [PubMed] [Google Scholar]
- 13.Choi H, Schoeni RF, Langa KM, Heisler MM. Older adults’ residential proximity to their children: Changes after cardiovascular events. J Gerontol B Psychol Sci Soc Sci. 2015;70:995–1004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Choi H, Schoeni RF, Langa KM, Heisler MM. Spouse and child availability for newly disabled older adults: Socioeconomic differences and potential role of residential proximity. J Geront B Pyschol Sci Soc Sci. 2015;70:462–469 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Saposnik G, Cote R, Rochon PA, Mamdani M, Liu Y, Raptis S, et al. Care and outcomes in patients with ischemic stroke with and without preexisting dementia. Neurology. 2011;77:1664–1673 [DOI] [PubMed] [Google Scholar]
- 16.Subic A, Cermakova P, Norrving B, Winblad B, von Euler M, Kramberger MG, et al. Management of acute ischaemic stroke in patients with dementia. J Intern Med. 2017;281:348–364 [DOI] [PubMed] [Google Scholar]
- 17.Jha AK, Orav EJ, Epstein AM. Low-quality, high-cost hospitals, mainly in south, care for sharply higher shares of elderly black, hispanic, and medicaid patients. Health Aff. 2011;30:1904–1911 [DOI] [PubMed] [Google Scholar]
- 18.Murao K, Bodenant M, Cordonnier C, Bombois S, Henon H, Pasquier F, et al. Does pre-existing cognitive impairment no-dementia influence the outcome of patients treated by intravenous thrombolysis for cerebral ischaemia? J Neurol Neurosurg Psychiatry. 2013;84:1412–1414 [DOI] [PubMed] [Google Scholar]
- 19.Luker JA, Wall K, Bernhardt J, Edwards I, Grimmer-Somers KA. Patients’ age as a determinant of care received following acute stroke: A systematic review. BMC Health Serv Res. 2011;11:161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chodosh J, Petitti DB, Elliott M, Hays RD, Crooks VC, Reuben DB, et al. Physician recognition of cognitive impairment: Evaluating the need for improvement. J Am Geriatr Soc. 2004;52:1051–1059 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.