Kidney Week

Abstract: TH-PO266

Wrist-Based Accelerometry and Risk of Emergency Department Visit or Hospitalization

Session Information

• Hemodialysis and Frequent Dialysis - II
  November 07, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Dialysis

• 701 Dialysis: Hemodialysis and Frequent Dialysis

Authors

• Munugoti, Samhitha, Albert Einstein College of Medicine, Long Island City, New York, United States

Samhitha Munugoti,

• Kotwani, Sonia, Albert Einstein College of Medicine, Long Island City, New York, United States

Sonia Kotwani,

• Randhawa, Lovepreet S., Albert Einstein College of Medicine, Long Island City, New York, United States

Lovepreet S. Randhawa,

• Dalezman, Solomon, Island Nephrology Services, Flushing, New York, United States

Solomon Dalezman,

• Elters, Antonio Carlos, Albert Einstein College of Medicine, Long Island City, New York, United States

Antonio Carlos Elters,

• Nam, Kate, Albert Einstein Medical School, Scarsdale, New York, United States

Kate Nam,

• Ibarra, Jose Sebastian, Albert Einstein College of Medicine, Long Island City, New York, United States

Jose Sebastian Ibarra,

• Venkataraman, Sandheep, Montefiore Medical Center, Bronx, New York, United States

Sandheep Venkataraman,

• Paredes, William, Albert Einstein College of Medicine, Long Island City, New York, United States

William Paredes,

• Ohri, Nitin, Albert Einstein College of Medicine, Long Island City, New York, United States

Nitin Ohri,

• Abramowitz, Matthew K., Albert Einstein College of Medicine, Long Island City, New York, United States

Matthew K. Abramowitz,

Background

Sedentary behavior has been associated with poor outcomes in dialysis patients. Whether changes in physical activity yield additional prognostic information is unknown.

Methods

We conducted a prospective study of wrist-based accelerometry in a single outpatient dialysis unit. 50 patients receiving thrice-weekly hemodialysis were enrolled and wore a commercial fitness tracker for 6 months which provided daily step count data. Information on emergency department (ED) visits and hospitalizations were obtained from monthly patient questionnaires, dialysis unit records, and our hospital electronic medical record. Each patient’s baseline activity level was defined as the mean step count during the 2 weeks following study entry. Poisson regression using generalized estimating equations and adjusted for age, sex, race, BMI, diabetes status, and follow-up time was conducted to examine the association of step counts with the composite outcome of ED visit or hospitalization, and with hospitalization alone.

Results

Data were excluded for 2 patients who wore the device <1 month. Daily step counts at baseline averaged 4538 +/- 3001. The mean age of the cohort was 60+/-13, 49% were women, 40% were black, 47% had diabetes, and the mean BMI was 28+/-7 kg/m². Compared with participants in the highest step count tertile, those in the lowest step count tertile were older but did not differ by other baseline characteristics. There were 38 occurrences of the composite outcome in 17 patients. After adjustment, each 1000-step reduction in daily step count was associated with 28% increased risk of both the composite outcome (incidence rate ratio (IRR) 1.28, 95% CI 1.07-1.53) and hospitalization alone (IRR 1.28, 95% CI 1.21-1.61). Each 10% reduction from an individual’s baseline activity level was associated with a 17% increased risk of the composite outcome (IRR: 1.17, 95%CI: 1.02-1.34).

Conclusion

Step count monitoring using a wearable device identifies changes in activity that are associated with increased risk of hospitalization or ED visit. Future studies should examine whether this approach could provide a real-time prediction of adverse events.

Funding

• NIDDK Support