Results differ by applying distinctive multiple imputation approaches on the longitudinal cardiovascular health study data

Recovery from Severe Disability that Develops Progressively Versus Catastrophically: Incidence, Risk Factors, and Intervening Events

BACKGROUND

Few prior studies have evaluated recovery after the onset of severe disability or have distinguished between the two subtypes of severe disability.

OBJECTIVES

To identify the risk factors and intervening illnesses and injuries (i.e., events) that are associated with reduced recovery after episodes of progressive and catastrophic severe disability.

DESIGN

Prospective longitudinal study of 754 nondisabled community-living persons, aged 70 years or older.

SETTING

Greater New Haven, CT, March 1998 to December 2016.

PARTICIPANTS

A total of 431 episodes of severe disability were evaluated from 385 participants: 116 progressive (115 participants) and 315 catastrophic (270 participants).

MEASUREMENTS

Candidate risk factors were assessed every 18 months. Functional status and exposure to intervening events leading to hospitalization, emergency department visit, or restricted activity were assessed each month. Severe disability was defined as the need for personal assistance with three or more of four essential activities of daily living. Recovery was defined as return to independent function (no disability) within 6 months of developing severe disability.

RESULTS

Recovery occurred among 35.3% (95% confidence interval [CI] = 26.0%-48.0%) and 61.6% (95% CI = 53.5%-70.9%) of the 116 progressive and 315 catastrophic severe disability episodes, respectively. In the multivariable analyses, lives alone, frailty, and intervening hospitalization were each independently associated with reduced recovery from progressive disability, with adjusted hazard ratios (95% CIs) of 0.31 (0.15-0.64), 0.23 (0.12-0.45), and 0.27 (0.08-0.95), respectively, whereas low functional self-efficacy, intervening restricted activity, and intervening hospitalization were each independently associated with reduced recovery from catastrophic disability, with adjusted hazard ratios (95% CIs) of 0.56 (0.40-0.81), 0.55 (0.35-0.85), and 0.45 (0.31-0.66), respectively.

CONCLUSIONS

Recovery of independent function is considerably more likely after the onset of catastrophic than progressive severe disability, the risk factors for reduced recovery differ between progressive and catastrophic severe disability, and subsequent exposure to intervening illnesses and injuries considerably diminishes the likelihood of recovery from both subtypes of severe disability.

Risk Factors and Precipitants of Severe Disability Among Community-Living Older Persons

IMPORTANCE

Severe disability greatly diminishes quality of life and often leads to a protracted period of long-term care or death, yet the processes underlying severe disability have not been fully evaluated.

OBJECTIVE

To evaluate potential risk factors and precipitants associated with severe disability that develops progressively (during ≥2 months) vs catastrophically (from 1 month to the next).

DESIGN, SETTING, AND PARTICIPANTS

Prospective cohort study conducted in greater New Haven, Connecticut, from March 1998 to December 2016, with 754 nondisabled community-living persons aged 70 years or older. Data analysis was conducted from November 2018 to May 2019.

MAIN OUTCOMES AND MEASURES

Candidate risk factors were assessed every 18 months. Functional status and potential precipitants, including illnesses or injuries leading to hospitalization, emergency department visit, or restricted activity, were assessed each month. Severe disability was defined as the need for personal assistance with at least 3 of 4 essential activities of daily living. The analysis was based on person-months within 18-month intervals.

RESULTS

The mean (SD) age for the 754 participants was 78.4 (5.3) years, 487 (64.6%) were women, and 683 (90.5%) were non-Hispanic white participants. The incidence of progressive and catastrophic severe disability was 3.5% and 9.7%, respectively, based on 3550 intervals. In multivariable analysis, 6 risk factors were independently associated with progressive disability (≥85 years: hazard ratio [HR], 1.6; 95% CI, 1.1-2.4; hearing impairment: HR, 1.7; 95% CI, 1.0-2.8; frailty: HR, 2.4; 95% CI, 1.6-3.7; cognitive impairment: HR, 2.0; 95% CI, 1.3-3.1; low functional self-efficacy: HR, 1.8; 95% CI, 1.2-2.8; low peak flow: HR, 1.7; 95% CI, 1.2-2.4), and 4 were independently associated with catastrophic disability (visual impairment: HR, 1.4; 95% CI, 1.1-1.8; hearing impairment: HR, 1.3; 95% CI, 1.0-1.7; poor physical performance: HR, 1.8; 95% CI, 1.3-2.5; low peak flow: HR, 1.3; 95% CI, 1.0-1.7). The associations of the precipitants were much more pronounced than those of the risk factors, with HRs as high as 321.4 (95% CI, 194.5-531.0) for hospitalization and catastrophic disability and 48.3 (95% CI, 31.0%-75.4%) for hospitalization and progressive disability. Elimination of an intervening hospitalization was associated with a decrease in the risk of progressive and catastrophic severe disability of 3.0% (95% CI, 3.0%-3.1%) and 12.3% (95% CI, 12.1%-12.5%), respectively. Risk differences were 0.6% (95% CI, 0.6%-0.6%) and 1.3% (95% CI, 1.3%-1.4%) for emergency department visit and 0.1% (95% CI, 0.1%-0.2%) and 0.4% (95% CI, 0.4%-0.4%) for restricted activity, and ranged from 0.1% (95% CI, 0.1%-0.1%) to 0.3% (95% CI, 0.3%-0.3%) for the independent risk factors, for progressive and catastrophic disability, respectively.

CONCLUSIONS AND RELEVANCE

The findings of this study suggest that whether it develops progressively or catastrophically, severe disability among older community-living adults arises most commonly in the setting of an intervening illness or injury. To reduce the burden of severe disability, more aggressive efforts will be needed to prevent and manage intervening illnesses or injuries and to facilitate recovery after these debilitating events.

Refining the Association Between Body Mass Index and Atrial Fibrillation: G-Formula and Restricted Mean Survival Times

Background

Previous studies assessing the association between body mass index (BMI) and atrial fibrillation (AF) did not account for time‐varying covariates, which may be affected by previous BMI. We illustrate how the g‐formula can account for time‐varying confounding.

Methods and Results

We included 4392 participants from the Framingham Heart Study who were AF free at ages 45 to 55 years, and followed them for up to 20 years. We estimated hazard ratios (HRs) comparing time‐varying nonobese versus obese with Cox models. We used the g‐formula to compare nonobese versus obese and 10% annual decrease in BMI (until normal weight is reached) versus natural course. We estimated HRs and differences in restricted mean survival times, the mean difference in time alive and AF free. We adjusted for sex, age, and time‐varying risk factors. Cox models indicated that nonobese participants had a decreased rate of AF versus obese participants (HR, 0.83; 95% CI, 0.72–0.97). G‐formula analyses comparing everyone had they been nonobese versus obese yielded stronger associations (HR, 0.73; 95% CI, 0.58–0.91). The restricted mean survival time was 19.22 years had everyone been nonobese and 19.03 years had everyone been obese (difference, 2.25 months; 95% CI, −0.66 to 5.16). When assessing a 10% annual decrease in BMI, the association was weaker (HR 0.96; 95% CI, 0.86–1.08).

Conclusions

Decreased BMI was associated with a lower rate of AF after accounting for time‐varying covariates that depend on previous exposure using the g‐formula, which Cox models cannot accommodate. Absolute measures like the restricted mean survival time difference offer context to relative measures of association.

The Factor Structure of the Boston Diagnostic Aphasia Examination, Third Edition

OBJECTIVE

The Boston Diagnostic Aphasia Examination (BDAE) is one of the most commonly used aphasia batteries. The newest edition has undergone significant revisions since its original publication in 1972, but existing evidence for its validity is lacking. We examined the construct validity of BDAE-3 and identified the factor structure of this battery.

METHOD

A total of 355 people with aphasia of various types and severity completed neuropsychological evaluations to assess their patterns of language impairment. A principal component analysis with varimax rotation was conducted to examine the components of BDAE-3 subtests.

RESULTS

Five components accounting for over 70% of the BDAE-3 total variance were found. The five language factors identified were auditory comprehension/ideomotor praxis, naming and reading, articulation-repetition, grammatical comprehension, and phonological processing.

CONCLUSIONS

Our results show that the BDAE-3 demonstrates good construct validity, and certain language functions remain primary, distinct language domains (i.e., receptive vs. expressive language) across severities of aphasia. Overall, our findings inform clinical practice by outlining the inherent structure of language abilities in people with aphasia. Clinicians can utilize the findings to select core BDAE-3 tests that are most representative of their respective functions, thereby reducing the total testing time while preserving diagnostic sensitivity. (JINS, 2019, 25, 772-776).

A statistical analysis protocol for the time-differentiated target temperature management after out-of-hospital cardiac arrest (TTH48) clinical trial

BACKGROUND

The TTH48 trial aims to determine whether prolonged duration (48 hours) of targeted temperature management (TTM) at 33 (±1) °C results in better neurological outcomes compared to standard duration (24 hours) after six months in comatose out-of-hospital cardiac arrest (OHCA) patients.

METHODS

TTH48 is an investigator-initiated, multicentre, assessor-blinded, randomised, controlled superiority trial of 24 and 48 hours of TTM at 33 (±1) ° C performed in 355 comatose OHCA patients aged 18 to 80 years who were admitted to ten intensive care units (ICUs) in six Northern European countries. The primary outcome of the study is the Cerebral Performance Category (CPC) score observed at six months after cardiac arrest. CPC scores of 1 and 2 are defined as good neurological outcomes, and CPC scores of 3, 4 and 5 are defined as poor neurological outcomes. The secondary outcomes are as follows: mortality within six months after cardiac arrest, CPC at hospital discharge, Glasgow Coma Scale (GCS) score on day 4, length of stay in ICU and at hospital and the presence of any adverse events such as cerebral, circulatory, respiratory, gastrointestinal, renal, metabolic measures, infection or bleeding. With the planned sample size, we have 80% power to detect a 15% improvement in good neurological outcomes at a two-sided statistical significance level of 5%.

DISCUSSION

We present a detailed statistical analysis protocol (SAP) that specifies how primary and secondary outcomes should be evaluated. We also predetermine covariates for adjusted analyses and pre-specify sub-groups for sensitivity analyses. This pre-planned SAP will reduce analysis bias and add validity to the findings of this trial on the effect of length of TTM on important clinical outcomes after cardiac arrest.

TRIAL REGISTRATION

ClinicalTrials.gov: NCT01689077 , 17 September 2012.

Time-limited home-care reablement services for maintaining and improving the functional independence of older adults

BACKGROUND

Reablement, also known as restorative care, is one possible approach to home-care services for older adults at risk of functional decline. Unlike traditional home-care services, reablement is frequently time-limited (usually six to 12 weeks) and aims to maximise independence by offering an intensive multidisciplinary, person-centred and goal-directed intervention.

OBJECTIVES

To assess the effects of time-limited home-care reablement services (up to 12 weeks) for maintaining and improving the functional independence of older adults (aged 65 years or more) when compared to usual home-care or wait-list control group.

SEARCH METHODS

We searched the following databases with no language restrictions during April to June 2015: the Cochrane Central Register of Controlled Trials (CENTRAL); MEDLINE (OvidSP); Embase (OvidSP); PsycINFO (OvidSP); ERIC; Sociological Abstracts; ProQuest Dissertations and Theses; CINAHL (EBSCOhost); SIGLE (OpenGrey); AgeLine and Social Care Online. We also searched the reference lists of relevant studies and reviews as well as contacting authors in the field.

SELECTION CRITERIA

We included randomised controlled trials (RCTs), cluster randomised or quasi-randomised trials of time-limited reablement services for older adults (aged 65 years or more) delivered in their home; and incorporated a usual home-care or wait-list control group.

DATA COLLECTION AND ANALYSIS

Two authors independently assessed studies for inclusion, extracted data, assessed the risk of bias of individual studies and considered quality of the evidence using GRADE. We contacted study authors for additional information where needed.

MAIN RESULTS

Two studies, comparing reablement with usual home-care services with 811 participants, met our eligibility criteria for inclusion; we also identified three potentially eligible studies, but findings were not yet available. One included study was conducted in Western Australia with 750 participants (mean age 82.29 years). The second study was conducted in Norway (61 participants; mean age 79 years).We are very uncertain as to the effects of reablement compared with usual care as the evidence was of very low quality for all of the outcomes reported. The main findings were as follows.Functional status: very low quality evidence suggested that reablement may be slightly more effective than usual care in improving function at nine to 12 months (lower scores reflect greater independence; standardised mean difference (SMD) -0.30; 95% confidence interval (CI) -0.53 to -0.06; 2 studies with 249 participants).Adverse events: reablement may make little or no difference to mortality at 12 months' follow-up (RR 0.97; 95% CI 0.74 to 1.29; 2 studies with 811 participants) or rates of unplanned hospital admission at 24 months (RR 0.94; 95% CI 0.85 to 1.03; 1 study with 750 participants).The very low quality evidence also means we are uncertain whether reablement may influence quality of life (SMD -0.23; 95% CI -0.48 to 0.02; 2 trials with 249 participants) or living arrangements (RR 0.92, 95% CI 0.62 to 1.34; 1 study with 750 participants) at time points up to 12 months. People receiving reablement may be slightly less likely to have been approved for a higher level of personal care than people receiving usual care over the 24 months' follow-up (RR 0.87; 95% CI 0.77 to 0.98; 1 trial, 750 participants). Similarly, although there may be a small reduction in total aggregated home and healthcare costs over the 24-month follow-up (reablement: AUD 19,888; usual care: AUD 22,757; 1 trial with 750 participants), we are uncertain about the size and importance of these effects as the results were based on very low quality evidence.Neither study reported user satisfaction with the service.

AUTHORS' CONCLUSIONS

There is considerable uncertainty regarding the effects of reablement as the evidence was of very low quality according to our GRADE ratings. Therefore, the effectiveness of reablement services cannot be supported or refuted until more robust evidence becomes available. There is an urgent need for high quality trials across different health and social care systems due to the increasingly high profile of reablement services in policy and practice in several countries.

A longitudinal, observational study with many repeated measures demonstrated improved precision of individual survival curves using Bayesian joint modeling of disability and survival

BACKGROUND/STUDY CONTEXT: It has not been previously demonstrated whether Bayesian joint modeling (BJM) of disability and survival can, under certain conditions, improve precision of individual survival curves.

METHODS

A longitudinal, observational study wherein 754 initially nondisabled community-dwelling adults in greater New Haven, Connecticut, were observed on a monthly basis for over 10 years.

RESULTS

In this study, BJM exploited many monthly observations to demonstrate, relative to a separate survival model with adjustment, improved precision of individual survival curves, permitting detection of significant differences between survival curves of two similar individuals. The gain in precision was lost when using only those observations from intervals of 6, 9, or 12 months.

CONCLUSION

When there are many repeated measures, BJM of longitudinal functional disability and interval-censored survival can potentially increase the precision of individual survival curves relative to those from a separate survival model. This may facilitate the identification of significant differences between individual survival curves, a useful result usually precluded by the large variability inherent to individual-level estimates from stand-alone survival models.

Functional Status and Social Contact Among Older Adults

This article tests whether functional status is associated with likelihood of social contact among older adults. Data come from the Second Longitudinal Study on Aging, a longitudinal nationally representative sample of 9,447 noninstitutionalized individuals aged 70 and over at baseline in 1995. Functional status is measured using an index of activities of daily living (ADL) and instrumental activities of daily living (IADL). Social contact is measured by asking respondents whether they had gotten together socially or talked on the phone with friends/neighbors or family in the past 2 weeks. Greater number of functional limitations is associated with a decreased likelihood of social contact at follow-up via the phone with friends (odd ratio [OR] = 0.94, p < .01) and family (OR = 0.96, p < .01), and a decreased likelihood of getting together with friends (OR = 0.93, p < .01) and family (OR = 0.97, p < .01). Results indicate that functional limitations have a broad impact on self-reported social contact among older adults.

Sensitivity analysis in multiple imputation in effectiveness studies of psychotherapy

The importance of preventing and treating incomplete data in effectiveness studies is nowadays emphasized. However, most of the publications focus on randomized clinical trials (RCT). One flexible technique for statistical inference with missing data is multiple imputation (MI). Since methods such as MI rely on the assumption of missing data being at random (MAR), a sensitivity analysis for testing the robustness against departures from this assumption is required. In this paper we present a sensitivity analysis technique based on posterior predictive checking, which takes into consideration the concept of clinical significance used in the evaluation of intra-individual changes. We demonstrate the possibilities this technique can offer with the example of irregular longitudinal data collected with the Outcome Questionnaire-45 (OQ-45) and the Helping Alliance Questionnaire (HAQ) in a sample of 260 outpatients. The sensitivity analysis can be used to (1) quantify the degree of bias introduced by missing not at random data (MNAR) in a worst reasonable case scenario, (2) compare the performance of different analysis methods for dealing with missing data, or (3) detect the influence of possible violations to the model assumptions (e.g., lack of normality). Moreover, our analysis showed that ratings from the patient's and therapist's version of the HAQ could significantly improve the predictive value of the routine outcome monitoring based on the OQ-45. Since analysis dropouts always occur, repeated measurements with the OQ-45 and the HAQ analyzed with MI are useful to improve the accuracy of outcome estimates in quality assurance assessments and non-randomized effectiveness studies in the field of outpatient psychotherapy.

Using multiple imputation to deal with missing data and attrition in longitudinal studies with repeated measures of patient-reported outcomes

Objective

Missing data is a ubiquitous problem in studies using patient-reported measures, decreasing sample sizes and causing possible bias. In longitudinal studies, special problems relate to attrition and death during follow-up. We describe a methodological approach for the use of multiple imputation (MI) to meet these challenges.

Methods

In a cohort of patients treated with percutaneous coronary intervention followed with use of repetitive questionnaires and information from national registers over 3 years, only 417 out of 1,726 patients had complete data on all measure points and covariates. We suggest strategies for use of MI and different methods for dealing with death along with sensitivity analysis of deviations from the assumption of missing at random, all with the use of standard statistical software. The Mental Component Summary from Short Form 12-item survey was used as an example.

Conclusion

Ignoring missing data may cause bias of unknown size and direction in longitudinal studies. We have illustrated that MI is a feasible method to try to deal with bias due to missing data in longitudinal studies, including attrition and nonresponse, and should be considered in combination with analysis of sensitivity in longitudinal studies. How to handle dropout due to death is still open for debate.

Respiratory impairment in older persons: when less means more

BACKGROUND

Among older persons, within the clinical context of respiratory symptoms and mobility, evidence suggests that improvements are warranted regarding the current approach for identifying respiratory impairment (ie, a reduction in pulmonary function).

METHODS

Among 3583 white participants aged 65 to 80 years (Cardiovascular Health Study), we calculated the prevalence of respiratory impairment using the current spirometric standard from the Global Initiative for Obstructive Lung Disease (GOLD) and an alternative spirometric approach termed "lambda-mu-sigma" (LMS). Results for GOLD- and LMS-defined respiratory impairment were evaluated for their (cross-sectional) association with respiratory symptoms and gait speed, and for the 5-year cumulative incidence probability of mobility disability.

RESULTS

The prevalence of respiratory impairment was 49.7% (1780/3583) when using the GOLD and 23.2% (831/3583) when using LMS. Differences in prevalence were most evident among participants who had no respiratory symptoms, with respiratory impairment classified more often by the GOLD (38.1% [326/855]) than LMS (12.3% [105/855]), as well as among participants who had normal gait speed, with respiratory impairment classified more often by the GOLD (46.4% [1003/2164]) than LMS (19.3% [417/2164]). Conversely, the 5-year cumulative incidence probability of mobility disability for respiratory impairment was higher for LMS than GOLD (0.313 and 0.249 for never-smokers, and 0.352 and 0.289 for ever-smokers, respectively), but was similar for normal spirometry by LMS or GOLD (0.193 and 0.185 for never-smokers, and 0.219 and 0.216 for ever-smokers, respectively).

CONCLUSIONS

Among older persons, the LMS approach (vs the GOLD approach) classifies respiratory impairment less frequently in those who are asymptomatic and is more strongly associated with mobility disability.