Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Ariz Geriatr Soc J. Author manuscript; available in PMC 2010 October 6.
Published in final edited form as:
Ariz Geriatr Soc J. 2008 May; 13(1): 15–18.
PMCID: PMC2950659

Quality Improvement in Nursing Homes: Identifying Depressed Residents is Critical to Improving Quality of Life

Neval L. Crogan, PhD, APRN, BC, GNP, FNGNA, Associate Professor
University of Arizona College of Nursing, Tucson, AZ
Bronwynne C. Evans, PhD, RN, CNS, FNGNA, Associate Professor


The prevalence of depression in nursing home residents is three to five times higher than in older adults from the community.1 Depression is thought to be related to the gloomy institutionalized environment and an assortment of losses, including those associated with function, independence, social roles, friends and relatives, and past leisure activities.2 Despite the public's increased awareness of depression, it remains underrecognized and undertreated by professionals who care for older residents in nursing homes.3 It seems intuitive that depression must be recognized before it can be treated, yet our national long-term care system continues to utilize an unreliable scale from the Minimum Data Set as its foundation for assessment. Warnings of the scale's inadequacy have been sounded repeatedly almost since its conception4,5 and its potential role in lack of recognition and treatment of depression by nursing home staff, nurse practitioners, and physicians is a troubling one.

The purpose of this article is to (1) report the prevalence of depression in a sub-sample of residents from a National Institutes of Health study whose depression was not detected by the MDS and, consequently, was previously untreated, (2) compare their nutritional and functional status with residents whose depressive states were previously detected by the MDS and treated, and (3) recommend quality improvement strategies for identification and treatment of depression in nursing home residents.

Assessment of Depression in Nursing Home Residents

In nursing homes, the MDS6 is a federally mandated, comprehensive assessment instrument that is completed within 14 days of admission, quarterly, and when a resident's condition changes. The Mood and Behavior Patterns section of the MDS assesses indicators of depression, anxiety, sad mood, mood persistence, and change in mood.6 Version 2 includes such items as “resident made negative statements, persistent anger and irritability, expressions of seemingly unrealistic fears, repetitive health complaints, repetitive anxious complaints, sad, pained, or worried facial expressions, and crying or tearfulness”. These items are scored on a scale from 0 (not exhibited in the preceding 30 days) to 2 (exhibited daily or almost daily).

Empirical data, however, continue to raise questions about the validity and reliability of the MDS in measuring depressive symptoms. Horgas and Margrett7 note that the MDS may substantially underestimate the prevalence of depression and that MDS scores are generally unrelated to medical diagnoses of depression or nursing assistants' ratings of resident depression, which tend to be congruent with one another. Their conclusions are consistent with those of other investigators5 who cautios that the MDS should not be used as a sole outcome measure of depressive symptoms because it does not adequately capture more subjective phenomena such as mood and depression.

Results from several other studies also call into question the accuracy of the MDS assessment. For example, Hendrix8 examined convergent validity of the Cornell Scale for Depression in Dementia (CSDS) in relation to the MDS and found that the MDS responded with a score of 0 (not exhibited) for both non-depressed (88–99%) and depressed residents (80–91%). These results are buttressed by two other studies in which the MDS missed 75% of major depressive disorders and 87% of all types of depression in nursing home residents.9,10 The MDS also did not perform well in another study when compared with the results of the Hamilton Depression Scale or the Geriatric Depression Scale (GDS), both well-validated instruments for depression assessment in older populations.1114

Lack of validity (whether or not it is the concept of depression which is actually being measured) may underlie lack of detection of depression by Versions 1 and 2 of the MDS, currently in use. The MDS may not clearly reflect affective status, particularly when symptoms overlap one another. It fails to differentiate, for example, among symptoms of pseudodementia (actually depression in older adults), dementia (gradual global decline in cognitive function), and delirium (acute, temporary disordered thinking and awareness, often caused by drug toxicity or dehydration). Complicating the process of assessment, some factors such as distractibility (a characteristic of depression) are addressed by the MDS as indicating delirium but this variable is not considered in the assessment of depression. Moreover, the MDS assesses current resident status, but does not predict future problems, so important potential problems are not identified or included in the plan of care.

Adding to the problem is the fact that some declines in physical health resemble the signs and symptoms of mood disorders, making clinical detection of depression problematic, especially if health care providers expect a more negative affect with age.13 Even nonverbal expressions change as the skin wrinkles and reaction times increase. Cultural and ethnic backgrounds that espouse different ways of expressing emotion and cohort influences such as the Great Depression may also confound assessment. For example, the current cohort of older adults who remember the 1930s believe strongly in self-sufficiency and view admissions of depression as unbecoming. Accordingly, they will often deny any affective symptoms.13

Differentiating depression from normal levels of sadness and worry in nursing home residents is not easy. Although extremes in affect are easy to distinguish, detecting subtle changes over time requires sensitive, specific instruments that have been validated for older populations (such as the GDS13,14). Use of such instruments results in more accurate assessments and care plans. In summary, the MDS alone may not provide an adequate assessment and. therefore, may delay or prevent detection of depression and early treatment. (The GDS was used for assessment of depressive symptoms in the study reported here.)


Research Design

This secondary data analysis was completed from a larger nutrition intervention study funded by the National Institutes of Health (NIH). A two-group prospective quasi-experimental design with measures taken at baseline and then at 6 months examined outcomes in residents who received a 6-month assessment process (including recommendations for improving nutrition; described below) and in residents from a comparison group receiving routine care specific to their nursing home. The purpose of the parent study was to test the feasibility of a new process of care (Individualized Nutrition Rx [INRx]) that utilizes a predictive model for identification of at-risk residents, a systematic assessment process, and evidence-based interventions to improve nutritional status.15,16 The INRx consists of the following components:

  1. Identification of at-risk residents (a part of the screening process).
  2. Resident choice of small-group dining or in-room dining.
  3. Comprehensive resident assessment by a facility interdisciplinary team consisting of a registered nurse, dietitian, and pharmacist.
    1. Individualized care planning and implementation based on risk factors and comprehensive assessment. Approaches may include adaptive equipment to enhance feeding ability; improved positioning; specific instructions for staff to cue residents to eat; or provision of finger foods so residents can feed themselves at their own pace.
    2. Organized referral to internal rehabilitation programs such as a progressive self-feeding program to relearn self-feeding skills, or to specialized rehabilitation (speech therapy, occupational therapy), dentistry, or residents' primary care physicians.
    3. Orientation of staff to implementation of new approaches.
  4. Bi-weekly re-assessment by registered nurse and dietitian; monthly re-assessment by pharmacist.
  5. Plan updates for care plan, referrals, and feedback to staff.

Sample and Setting

Eighty-one residents were recruited to participate in the NIH study. Of those 81, 40 were from 2 southwest nursing homes that served as comparison sites. The remaining 41 were residents at a third intervention nursing home. Of the 41 intervention residents, 12 were already undergoing treatment for depressive symptoms detected by nursing staff using the MDS. Another 8 residents exhibited evidence of depression when assessed at baseline with the GDS by the research team. These 20 residents with depression are the focus of this secondary data analysis. All procedures were approved by the University of Arizona Human Subjects Review Board prior to initiation of the NIH study. Written informed consent was obtained from each participating resident or their legal guardians.

Study Measures

Depression was measured using the Geriatric Depression Scale,17,18 a valid indicator of depression in mildly to moderately cognitively impaired elders.19 The GDS is a 30-item questionnaire that asked participants to respond yes or no about how they felt on the day of administration. Items were summed to obtain a total score. Scores of 0–10 are considered normal, 11–20 indicates moderate depression, and 21–30 indicates severe depression. The scale has been shown to have internal consistency and it correlates highly with other established depression scales (e.g. Hamilton Rating Scale for Depression, Zung Self Rating Scale for Depression).

Body Mass index (BMIj is determined by dividing the resident's weight in kilograms by the height in meters squared. Both weights and heights were measured by a research nurse. Weights were obtained at rising while the resident was still dressed in bed clothes, on a chair scale that was calibrated weekly by maintenance staff. Knee height (length from sole of foot to anterior thigh with both ankle and knee bent at 90-degree angle) was measured and height was calculated, based on gender specific formulas.20 This procedure avoided inaccuracy associated with standing height due to loss of vertebral mineralization and volume in intervertebral disks. BMI was calculated from these two figures.

Serum prealbumin is an indicator of protein status and dietary protein intake (normal 10–40 mg/dL). A local laboratory company provided phlebotomy services and laboratory analysis using standard testing methods for albumin (Bromcresol green testing method) and prealbumin (Photometric turbid metric testing method). Both of these methods have good sensitivity and reproducibility.21,22

Functional status was measured using the 6-item Katz Index of Independence in Activities of Daily Living (Katz ADL). The Katz ADL measures what residents actually do rather than what the person is capable of doing and has consistently demonstrated its utility in older populations. Clients are scored yes or no for independence in bathing, dressing, toileting, transferring, continence, feeding. A score of six indicates full function; four indicates moderate impairment, and two or less indicates severe functional impairment. Overall, ADL scores correlate with range of motion and cognitive function.23 The Katz ADL is highly reliable and sensitive to change in elderly in homecare and boarding homes.24 The coefficient of reproducibility (which measures Guttman characteristics) is 0.95 – 0.98. Tests of validity show the scale is highly correlated with house confinement, mobility post-hospital, and with the MMSE (r = 0.76).23,25,26

Data Management and Analysis

A research associate entered all data from study measures into a computer database. All data were cleaned and organized prior to analyses. Resident groups were compared on all demographic and study variables at baseline to ensure initial comparability using t tests for interval data, Chi square for nominal data and the Mann-Whitney U statistic for ordinal data. Outcome data were analyzed using repeated measures ANOVA. The dependent variables were nutritional status (BMI, prealbumin), depression (GDS), and functional status (Katz ADL Index).


During initial recruitment and screening, 20 of 41 (49%) intervention residents demonstrated symptoms of depression. Nineteen of the depressed residents were Anglo (one was of Hispanic origin) and one African-American. Sixty-five percent (n=13) were female, 35% were male (n=7). Of those 20 residents with symptoms of depression, only 60% (n=12) of the cases were detected by nursing staff using the MDS. In the remaining 8 residents (40%), depression had not been detected; consequently, these residents were not receiving treatment prior to the study. Of note, those not identified by staff as depressed were, on average, older than those whose depression was detected (86.88 ± 4.76 vs. 81.58 ± 9.44), perhaps reflecting increased difficulty in assessment of depression with advancing age.

The eight residents with previously undetected depression scored on average 17 on the GDS, indicating moderate depression (Table 1). These scores were similar to the 12 residents who had already been diagnosed and receiving treatment. Post INRx intervention, both groups had clinically, although not statistically, significant improvements in GDS scores. In fact, the percentage of residents with depression decreased 17.1% post intervention (46.3% to 29.2%).

Table 1
Mean and standard deviation differences between depressed residents detected and undetected at baseline and post intervention (6 months later)

Even though protein stores (prealbumin) improved post intervention in all 20 residents, the baseline prealbumin range was more variable in the previously undetected group (range 10–50 mg/dL; normal 10–40 mg/d). Two of these eight residents had abnormal prealbumin results; one resident had a prealbumin of 10 mg/dL, indicating moderate protein depletion,27 and another had a prealbumin of 50 mg/dL, possibly due to either alcohol dependence or an adverse drug event.28


Although depression is a treatable illness, the research team found that 8 of 20 depressed (40%) intervention residents did not have a treatment plan to address depression, indicating that facility staff failed to detect depressive symptoms in these at-risk nursing home residents. These results underscore concerns on the part of residents, nursing home staff, physicians, and federal regulators about the prevalence of depression in nursing homes. For example, the Centers for Medicare and Medicaid Services (CMS) has incorporated the “Percent of Residents who are More Depressed or Anxious” as one of 15 nursing home quality measures.29 These quality measures reflect resident assessment (MDS) data that nursing homes routinely collect, and they are intended to be used by consumers to compare nursing homes, and by facility staff to help focus quality improvement efforts. However, the fundamental problem of inaccurate or incomplete assessment renders them less than useful in these endeavors.

Attentive nursing home staff is the key to detection and early treatment of depression. In nursing homes, however, nursing assistants provide most of the hands-on care for residents, while nurses give medications and treatments but only oversee most bedside care. Nursing assistants, whose capabilities are often viewed somewhat skeptically, can provide valuable data that supports detection of depression,7 and should be trained to recognize and report signs and symptoms of depression. Although nursing assistants do not traditionally receive extensive training or practice in rating behaviors indicative of depression, one study using the Revised Memory and Behavior Problems Scale (RMBPS) checklist to assess both depression and dementia in 135 nursing home residents found that nursing assistants' ratings of resident depression were more congruent with medical diagnoses than ratings done by nurses using the MDS.7 These investigators note that their findings may reflect the fact that nursing assistants provide 80–90% of direct resident care and may be more closely attuned to residents' actual behaviors than nurses. Thus, a facility-wide quality improvement initiative (Figure 1) focused on (1) training nursing assistants to consistently and accurately report targeted observations to nurses, and on (2) educating nurses about validated measures for assessment of depression, is critical to detection and treatment.

Figure 1
Quality Improvement Initiative for Nursing Homes

In this study, depression screening using the GDS resulted In cosiderable numbers of residents receiving treatment for Depression, congruent with the results described by other investigators.30 Of note, both newly treated and those who were already being treated for depression on entrance to the study improved their scores on the GDS over the course of the six-month intervention.

The increase demonstrated in all 20 residents' protein stores over the course of the six-month study could have reflected the initiation, and continuation, of treatment for depression, with subsequent increased appetite. These increased protein stores occurred even in the face of severe functional impairment in both detected and undetected residents at baseline, which remained so even after the intervention. Functional impairment is positively correlated with poor food intake, nutritional deficiencies, and depression.3133

Early assessment of depression and use of anti-depressant medications could automatically address depression as a risk factor for malnutrition among nursing home residents and have a positive effect on food intake and prevention of nutritional deficiencies.2,34,35 In this portion of the NIH study with 20 residents, the percentage with depression decreased 17.1% during the six-month intervention (from 46.3% to 29.2%) while the percentage of comparison nursing home residents with depression decreased only 9.4% (from 51.3% to 41.9%). Concurrently, protein stores increased in both previously detected and previously undetected residents with depression. Increases in protein stores were greater in residents whose depression was previously detected, perhaps signaling the positive effects of a longer duration of treatment.

Without treatment, the incidence of depression leads to adverse sequelae such as functional impairment and increased morbidity and mortality36 which influences health care utilization and drives costs up. Appraisals of the cost of depression probably underestimate the economic burden of the disease, because costs borne by caregivers and family members and the cost of complications (such as nutritional deficiencies and resultant pressure ulcers in nursing home residents) from lack of treatment are difficult to compute.37 However, in the general population, the economic burden of depression was about $83 billion in 2000, of which $26 billion were direct treatment costs and $5 billion were suicide-related costs.37 Undoubtedly, timely intervention and treatment is preferred and is, over the long term, more humane and cost effective. Until the draft of the MDS 3, which uses more traditionally recognized items to measure depression such as “feeling bad about yourself”, is completed and tested to ensure adequate psychometrics, healthcare providers must supplement the MDS with more effective instruments such as the GDS.


As shown in this study, using the MDS to detect depression in older nursing home residents “is of limited clinical value.”11(p. 437) Depression is rampant in nursing homes because of stressful losses such as loss of autonomy, privacy, functional status, a friend or family member; alcohol or substance abuse; and a history of previous depressive episodes.10 However, depression is a treatable illness and nursing home staff are in the forefront of the detection effort. A quality improvement initiative focused on training nursing assistants to identify and report targeted observations to nurses, and nurses who use validated assessment measures for depression could make a critical difference in detection and treatment of this problem. Such measures take little additional staff time to administer but have large pay-offs in resident quality of life. Additionally, effective screening can serve as the impetus for social services and primary care physicians to refer residents to psychiatrists for further evaluation and treatment.


Funding source: NIH/NINR 1 R15 NR008382-01A1


1. NIH Consensus Development Panel on depression in late life Diagnosis and treatment of depression in late life. JAMA. 1992;268:1018–1024. [PubMed]
2. Fitzsimmons S. Easy rider wheelchair biking: A nursing-recreation therapy clinical trial for the treatment of depression. Journal of Gerontological Nursing. 2001;27(5):14–23. [PubMed]
3. Brown MN, Lapane KL, Luisi AF. The management of depression in older nursing home residents. JAGS. 2002;50:69–76. [PubMed]
4. Casten R, Lawton M, Parmelee PA, et al. Psychometric characteristics of the Minimum Data Set I: Confirmatory factor analysis. Journal of the American Geriatrics Society. 1998;46(6):726–735. [PubMed]
5. Teresi J, Holmes D. Should MDS data be used for research? The Gerontologist. 1992;32:148–151. [PubMed]
6. American Health Care Association . Facts and trends, the nursing facility sourcebook. American Health Care Association; Washington, DC: 1998.
7. Horgas A, Margrett J. Measuring behavioral and mood disruptions in nursing home residents using the Minimum Data Set. Outcomes Management for Nursing Practice. 2001;5(1):28–32. [PubMed]
8. Hendrix C. Doctoral dissertation. Louisiana State University Health Sciences Center School of Nursing; 2001. Detecting patterns of depression using items in the indicators of depression, anxiety, and sad mood in the Minimum Data Set 2.0 among elderly residents in selected nursing home settings.
9. Cohen C, Hyland K, Framowitz I. Depression among nursing home residents with dementia: Final Report. New York State Department of Health. Bureau of Long-Term Care Services; Albany: 1995.
10. Cohen C, Hyland K, Magai C. Depression among African-American nursing home patients with dementia. American Journal of Geriatric Psychiatry. 1998;6:162–175. [PubMed]
11. Anderson R, Buckwalter K, Buchanan R, et al. Validity and reliability of the Minimum Data Set Depression Rating Scale (MDSDRS) for older adults in nursing homes. Age and Ageing. 2003;32:435–438. [PubMed]
12. McCurren C. Assessment for depression among nursing home elders: Evaluation of the MDS Mood Assessment. Geriatric Nursing. 2002;23(2):103–108. [PubMed]
13. Kane RL, Kane RA. Assessing older persons. Oxford University Press; New York: 2002.
14. Meeks S. Further evaluation of the MDS Depression Scale versus the Geriatric Depression Scale among nursing home residents. Journal of Mental Health in Aging. 2004;10(4):325–335.
15. Crogan NL, Alvine C. Testing of the Individual Nutrition Rx assessment process among nursing home residents. Applied Nursing Research. 2006;19:102–104. [PMC free article] [PubMed]
16. Crogan NL, Alvne C, Pasvogel A. Improving nutrition care for nursing home residents using the INRx process. Journal of Nutrition for the Elderly. 2007;25(3/4):XX. [PMC free article] [PubMed]
17. Sheikh RL, Yesavage JA. Geriatric Depression Scale (GDS) recent evidence and development of a short version. Clinical Gerontologist. 1986;5:165–173.
18. Yesavage JA, Brink TL, Rose TL, et al. Development and validalion of a geriatric depression screening scale: A preliminary report. Journal of Psychiatric Research. 1983;17:37–49. [PubMed]
19. Parmelee PA, Katz IR, Lawton MP. Depression among institutionalized aged: Assessment and prevalence estimation. Journal of Gerontology: Medical Sciences. 1989;44(1):M22–29. [PubMed]
20. Moore MC. Nutritional care. Mosby, Inc; St. Louis: 2001.
21. Sonora Quest Laboratories . Computer report of laboratory tests. Sonora Quest Laboratories of Tucson; AZ: 2002.
22. Henry JB. Clinical diagnosis and management by laboratory methods. W.B.S.; New York: 2001.
23. Katz S, Down TD, Cash HR, et al. Process in the development of the index of ADL. Gerontologist. 1970;10:20–30. [PubMed]
24. Sherwood SJ, Morris J, Mor V, et al. Compendium of measures for describing and assessing long-term care populations. Hebrew Rehabilitation Center for the Aged; Boston: 1977.
25. Frank-Stromborg M, Olsen SJ. Instruments for clinical health-care research. Jones and Bartlett Publishers; Sudbury, MA: 1997.
26. Katz S. Assessing self-maintenance: Activities of daily living, mobility and instrumental activities of daily living. Journal of the American Geriatrics Society. 1983;31(12):721–726. [PubMed]
27. Fischbach F. A manual of laboratory and diagnostic tests. 7th ed. Lippincott Williams & Wilkins; New York: 2004.
28. Epocrates PDA program. accessed at
29. Centers for Medicare and Medicaid Services Nursing Home Quality Initiatives. accessed at: [PubMed]
30. Cohen C, Hyland K, Kimhy D. The utility of mandatory depression screening of dementia patients in nursing homes. American Journal of Geriatric Psychiatry. 2003;160(11):2012–2017. [PubMed]
31. Crogan NL, Corbett C. Predicting malnutrition in nursing home residents using the Minimum Data Set. Geriatric Nursing. 2002;23(4):224–226. [PubMed]
32. Crogan NL, Corbett C, Short R. The Minimum Data Set: Predicting malnutrition in newly admitted nursing home residents. Clinical Nursing Research. 2002;11(3):341–353. [PubMed]
33. Corbett C, Crogan NL, Short R. Using the Minimum Data Set to predict weight loss in nursing home residents. Applied Nursing Research. 2002;15(4):249–253. [PubMed]
34. Schlenker ED. Nutrition in aging. 3rd Ed. McGraw-Hill; New York, NY: 1998.
35. Shils ME, Olson JA, Shike M. Modern nutrition in health and disease. 8th Ed. Lea & Febiger; Philadelphia: 1994.
36. Kane RL, Ouslander JG, Abrass IB. Essentials of clinical geriatrics. 5th ed. McGraw Hill Companies; New York: 2004.
37. Greenberg P, Kessler R, Birnbaum H, et al. The economic burden of depresson in the United States: how did it change between 1990 and 2000? Journal of Clinical Psychiatry. 2003;64(12):1465–75. [PubMed]