Clinical momentum in the care of older adults with advanced dementia: What evidence is there in the medical record?

Abstract

Overtreatment at the end of life contributes to poor quality of life, is often discordant with patient preferences, and strains healthcare systems.^{1, 2} “Clinical momentum” is a conceptual model to describe the latent, systems-level forces that create an inevitable trajectory toward intervention near the end of life.³ Feeding tube placement in patients with advanced dementia is a clear example of overtreatment at the end of life, given the associated harms and limited benefits.⁴ This study builds on qualitative work describing the contribution of clinical momentum to feeding tube placement and aims to identify previously characterized markers of clinical momentum in the medical record.⁵

We conducted a retrospective, single-center, matched case-control study. We used an Electronic Health Record (EHR) search to identify all hospitalized older adults (age ≥65) with a dementia diagnosis and activated healthcare agent during an unplanned admission of ≥3 days between January 2015 and December 2022. We confirmed patients' dementia diagnosis via chart review. The case group included patients who received a permanent feeding tube (i.e., Gastrostomy, Gastrojejunostomy, or Jejunostomy tube). We 1-1 matched controls, patients who did not receive a permanent feeding tube, with cases based on age, sex, and admitting diagnosis. We excluded patients with a preexisting feeding tube. The University of Wisconsin institutional review board deemed this study exempt.

We abstracted data from one admission per patient, capturing controls' last encounter, using a standardized manual chart abstraction form, including hospital events (e.g., aspiration, consultations) identified in previous qualitative work.⁵ We used Fisher's exact tests and t-tests to compare groups' baseline characteristics. We cataloged hospital trajectories and the frequency of event combinations leading to permanent feeding tube placement in the case group and death or discharge for the control group, displayed via UpSet Plot.⁶ Analyses were performed in SAS software (version 9.4, SAS Institute Inc., Cary, North Carolina).

We identified 34 cases and 34 matched controls. The mean age (SD) was 80.2 (8.7), and 34 (50%) were male. Demographic characteristics were similar between groups (Table 1). Although not statistically significant, case patients had a lower mean Charlson comorbidity score (8.8 (3.6) vs. 10.3 (3.2), p = 0.082).

The median (interquartile range (IQR)) length of stay was 21.5 (15–43) days among cases and 5 (4–9) days among controls. At the median, feeding tubes were placed on day 15 (IQR 7–18) of admission. One case patient and three control patients died in the hospital. On average, patients in the case group experienced 9.0 (2.2) unique hospital events before feeding tube placement, while control patients experienced, on average, 4.9 (2.5) events during their admission. We found higher rates for all events among case patients (Figure 1). Notably, 18 (52.9%) cases and seven (20.6%) controls had a “goals of care” meeting, while 10 (29.4%) cases and four (11.8%) controls received a palliative care consult. Twenty-seven (79.4%) cases and nine (26.5%) controls received a bedside swallow consult, diet modification, nutrition consult, and swallow therapy, the most common combination for both groups (Figure 1). Twenty-five (73.5%) cases and five (14.7%) controls also received a nasogastric tube (NG/DHT) with this combination.

Patients with advanced dementia who received a permanent feeding tube spent more time in the hospital, had more events, and were more likely to have received a nasogastric tube despite having more goals of care conversations, palliative care consultations, and fewer baseline comorbid conditions. Our previously reported qualitative study showed behavioral components of clinical momentum, for example, recognition primed decision-making and sunk costs, drive clinicians to prioritize a biomedical fix that initiates a care trajectory that is difficult to disrupt.⁵ In this light, these results suggest that prolonged exposure to a system designed to intervene on isolated problems can produce guideline-discordant care. Momentum increases the longer patients stay in the system, leading to additional tests and interventions that are difficult to withdraw once they are in place. For example, clinicians place a temporary nasogastric tube to address a problem, like aspiration, leading to a permanent tube when reinforced by downstream forces, for example, nursing home rules require patients have a permanent tube. Less prognostic uncertainty, resulting from more comorbid conditions among control patients, may facilitate earlier recognition that usual patterns of care are inappropriate.

This study is limited by the small nondiverse sample and the EHR data source, making it difficult to ascertain dementia diagnosis and stage due to coding variations,⁷ and limiting our ability to understand context, unique situations, and patients' acute illness severity. Future efforts to measure clinical momentum at scale may reveal novel targets to improve end-of-life care.

Study concept and design: MLS; Acquisition of subjects and/or data: KJB, TB, AB, KZ, AD, SIZ, MF; Analysis and interpretation of data: LNS, BMH, MLS; Preparation of manuscript: LNS, MLS, BMH, KJB, TB, AB, KZ, AD, SIZ, MF, KK.

The authors have no conflicts.

This study is funded by the National Institutes of Health (1R21AG068720-01). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

This study is funded by the National Institutes of Health (1R21AG068720-01). This manuscript has not been submitted or presented elsewhere.