Clinical research study
Physical symptoms as a predictor of health care use and mortality among older adults

https://doi.org/10.1016/j.amjmed.2004.12.015Get rights and content

Purpose

To describe the patterns of physical symptoms in older adults and to examine the validity of symptoms in predicting hospitalization and mortality.

Subjects and methods

Adults aged 60 years and older (N=3498) who completed screening for self-reported symptoms at routine primary care visits. Self-reported symptoms were collected using an abbreviated PRIME-MD screening instrument. Clinical characteristics, hospitalization, and mortality in the year following screening were measured using data taken from a comprehensive electronic medical record.

Results

The mean patient age was 69 years, 69% were women, and 56% were African-American. A majority (51%) of respondents characterized their health as fair or poor. The most commonly reported symptoms were musculoskeletal pain (65%), fatigue (55%), back pain (45%), shortness of breath (41%), and difficulty sleeping (38%). A summary score of physical symptoms (range 0–12) was a significant independent predictor of future hospitalization and death even when controlling for clinical characteristics, chronic medical conditions, self-rated health, and affective symptoms. Disease-specific symptoms were more common among patients diagnosed with the specific condition but there was also a substantial background prevalence of these symptoms.

Conclusion

Physical symptoms are highly prevalent in older primary care patients and predict hospitalization and mortality at one year. Future work is needed to determine how to target symptoms as a potential mechanism to reduce health care use and mortality.

Introduction

The prevalence and patterns of physical symptoms among older adults and their impact on health care utilization are areas of limited research. Symptoms are the leading reason that patients seek medical care and frequently serve as the basis for establishing a diagnosis.1, 2 Patients’ self-reports of symptoms also guide treatment decisions because they serve as an indicator of the severity of illness and the response to therapy. However, at least one-third of symptoms cannot be adequately explained even after medical evaluation, and the cost-effectiveness of diagnostic and management strategies are not well-defined for many common symptoms.2, 3 Attribution of specific symptoms to a single disease becomes increasingly problematic among older adult patients because of comorbid conditions.

Prior studies among older adults have described a high prevalence of symptoms but an unclear relationship between symptoms and help-seeking behavior.4, 5, 6 For example, Hale et al. found that 94% of older adults attending an outpatient clinic reported at least one symptom.5 However, Brody et al. reported that only 1% of all physical symptoms experienced by older patients were reported to physicians.4 Because older patients suffer a higher disease burden, it may seem self-evident that they would experience more symptoms, but the evidence is inconclusive.7, 8, 9, 10, 11, 12, 13, 14 Also, one might assume that symptoms not reported to a physician are inconsequential or not an indicator of serious disease.15 Unfortunately, for a whole host of serious conditions, such as myocardial infarction, depression, and dementia, early but important symptoms often go unreported.

Measures of physical function or disease burden have been used to identify older adults at high-risk for excess health care utilization, disability, or mortality. We hypothesized that patient’s self-reported symptoms would be an independent predictor of subsequent health services use and mortality even when controlling for diagnosed chronic conditions. The purpose of this study is three-fold. First, we describe the prevalence of common symptoms among older adults presenting to an urban primary care clinic. Second, we explore the predictive validity of symptoms in describing future health care utilization and mortality. Third, we test the strength of this relationship in multivariable models controlling for demographic and clinical characteristics.

Section snippets

Methods

This study was approved by the Indiana University Purdue University—Indianapolis Institutional Review Board. Between July 1999 and August 2001, all patients ≥ 60 years of age attending a large urban primary care clinic were administered a modified version of the PRIME-MD as a screen for common symptoms.16 The impetus of the screening program was a multi-center study of late life depression.17 Patients were approached at the time of clinic visits and asked to complete the PRIME-MD to assess

Prevalence of symptoms and their predictive validity across total sample

Over a 2-year period, 3675 older adults were approached and 3498 (95%) completed the modified PRIME-MD; 172 screens were refused (2.8%) and 5 were incomplete. The patient sample is two-thirds women and over half are African-American. Consistent with the high prevalence of chronic conditions, 51% of the subjects rate their health as fair or poor (Table 1). In addition, 80% of this cohort was economically disadvantaged as defined by federal or county health insurance assistance.

The most commonly

Discussion

We have described the prevalence of self-reported somatic symptoms among a large cohort of vulnerable older adults cared for in an urban primary care clinic. These findings demonstrate a high frequency of individual symptoms and a heavy burden of total symptoms among this group of older adults. Total symptom count was significantly correlated with hospitalization and death over the subsequent year even when controlling for demographic characteristics, physician-diagnosed chronic medical

Acknowledgments

Dr. Callahan was supported by a Career Leadership Award from the National Institute on Aging (K07 AG00868). The study was supported in part by grants from the John A. Hartford Foundation, Inc., and the Marion County Health Department.

References (30)

  • E.C. Hammond

    Some preliminary findings on physical complaints from a prospective study of 1,064,004 men and women

    Am J Public Health.

    (1964)
  • J.I. Escobar et al.

    Somatization in the community

    Arch Gen Psychiatry.

    (1987)
  • K. Kroenke et al.

    The prevalence of symptoms in medical outpatients and the adequacy of therapy

    Arch Intern Med.

    (1990)
  • G. Tibblin et al.

    Symptoms by age and sex

    Scand J Prim Health Care.

    (1990)
  • K. Kroenke et al.

    Symptoms in the communityprevalence, classification, and psychiatric comorbidity

    Arch Intern Med.

    (1993)
  • Cited by (0)

    View full text