At the time this article was submitted, Ms. Toth was a geriatric nurse practitioner student at the University of Minnesota School of Nursing, Minneapolis, Minnesota. Dr. Bliss is Professor in Long-Term Care of Elders and Interim Associate Dean for Research, Ms. Savik is Senior Statistician, and Dr. Wyman is Professor and Cora Miedl Siehl Endowed Chair in Nursing Research, University of Minnesota School of Nursing, Minneapolis, Minnesota. Dr. Wyman is also Professor of Community Health and Family Practice, School of Medicine, University of Minnesota.
The authors thank the staff and administration of the nursing homes for their cooperation with this study. They acknowledge data collection by Lisa Rice and Jill Scholz, who were graduate research assistants at the University of Minnesota School of Nursing.
This study was funded in part by a grant from the Gerontological Nursing Interventions Research Center NIH #P30 NR03970 [PI: Toni Tripp-Reimer, The University of Iowa College of Nursing] and an unrestricted faculty development grant by 3M Medical Division, both to Dr. Donna Z. Bliss.
Address correspondence to Anna M. Toth, RN, BAN, MSN, 517 Deer Trail, Montgomery, MN 56069; e-mail: firstname.lastname@example.org.
Perineal dermatitis (PD) is an inflammation of the skin in the genital, buttock, and thigh areas that can result in decreased well-being and increased health care costs. In mild cases, PD manifests as redness and discomfort. In more severe cases, there is loss of superficial skin layers, a rash or oozing vesicles, and pain. Secondary fungal infection may also occur (Brown & Sears, 1993). It is estimated that 35% of older adults who are incontinent have a rash or redness requiring treatment; the estimated annual cost of such skin irritations in nursing home residents with urinary incontinence is $6.6 billion (Wilson, Brown, Shin, Luc, & Subak, 2001). Yet, few studies about PD or its associated factors in older adults in any setting have been conducted.
Bliss, Savik, Harms, Fan, and Wyman (2006) investigated the prevalence and correlates of PD in nursing home residents, guided by a conceptual model of PD (Brown & Sears, 1993). Minimum Data Set (MDS) items were used to represent the factors in the model. Because numerous items from the MDS could be used to define each PD factor, composite variables of MDS items needed to be developed (Savik, Fan, Bliss, & Harms, 2005). This study reports the preliminary analysis of the validity of the MDS items used to operationally define those PD factors. Determining the validity of sources of data or measures of variables is important to show the degree to which they measure what they intend to measure. Criterion validity is determined by comparing the relationship between a data source and an external criterion (Polit & Beck, 2004). When the data obtained within the instruments agree or correlate highly with those of the external criterion, the instrument is validated.
The purpose of this study was to ascertain the validity of items on the MDS that are postulated to be associated with PD. This was done by determining the agreement between data on the MDS and in the medical and nursing records of nursing home residents. Determining the validity of the MDS items was an important step before the PD factors they defined could be used in any subsequent analysis (Bliss et al., 2006).
The MDS is a standardized instrument used for the comprehensive assessment and care planning of nursing home residents; it provides a rich source of information that is useful for researching the health problems of nursing home residents. Federal law mandates that all residents in nursing facilities certified by Medicare or Medicaid be assessed using the MDS at admission, with any significant change in health condition, and at set intervals (Centers for Medicare & Medicaid Services, 2006).
Several studies have reported the validity of the MDS overall (Casten, Lawton, Parmelee, & Kleban, 1998; Lawton, Casten, Parmelee, Van Haitsma, & Kleban, 1998; Morris, Jones, Fries, & Hirdes, 2004). Previous studies have also validated selected items on the MDS by comparing them with other validated scales and nursing home chart records. For example, the validity of MDS items related to mood (Meeks, 2004), pain (Cohen-Mansfield, 2004), and cognition (Cohen-Mansfield, Taylor, McConnell, & Horton, 1998; Gruber-Baldini, Zimmerman, Mortimore, & Magaziner, 2000) showed fair to good correlations that were statistically significant; MDS items were also found to be predictive of protein calorie malnutrition (Crogan & Corbett, 2002).
The validity of some MDS items can differ depending on the sample. Therefore, validating MDS data is important when using it for research.
When compared with nursing home chart data, MDS data related to falls collected over a longer period (180 days) showed good agreement (kappa = 0.50, p < 0.001); shorter term (collected in 30 days) MDS data related to falls had fair agreement with chart data (kappa = 0.29, p < 0.001) (Hill-Westmoreland & Gruber-Baldini, 2005). Chart documentation of incontinence care processes was consistent with MDS reporting; for example, MDS data about residents’ requiring toileting assistance agreed with chart documentation about residents’ reports of needing and having received toileting assistance. However, some differences were found when MDS items were compared with residents’ recall of care instead of the nursing home chart data (Schnelle et al., 2003). The validity of some MDS items can differ depending on the sample. Therefore, validating MDS data is important when using it for research. Using nursing home residents’ charts as a criterion against which to compare MDS data is a standard approach for validating the MDS.
© 2008/DigitalVision/Getty Images
Records of residents in two nonprofit nursing homes in the Twin Cities area of Minnesota were analyzed. Each nursing home housed residents on three floors; a total of 106 residents lived in one home and 184 lived in the other. Residents were eligible for the study if they had a full MDS assessment within the previous 12 months, were age 65 or older, and experienced urinary or fecal incontinence. A nursing home staff member used MDS data to screen all residents for meeting the inclusion criteria, and other nursing home staff confirmed eligibility.
Given the intensity of the chart review process, time involved in contacting some of the residents’ legal representatives for informed consent, and amount of funds available for the study, a sample size of approximately 40 residents (approximately 20 from each nursing home) was planned. Eligible residents were stratified by kind of incontinence, location in the nursing home, and whether they had perineal skin damage. Because it is not possible to determine using the MDS whether skin damage is in the perineal area, this information was obtained from residents’ chart records.
Using a list of computer-generated random numbers, residents were selected for recruitment from each stratum on the basis of their proportion in the nursing home. Random recruitment from strata was done to avoid inclusion of residents with only one kind of incontinence or skin status or from one floor. The number to be recruited from each stratum was determined by the proportion of residents in that stratum to the entire nursing home population. Residents who met the inclusion criteria were asked by nursing home staff whether they were interested in discussing the study with the investigators. If residents were unable to give consent, their legal representative was contacted in person or by letter by the nursing home staff. Research assistants (RAs) (graduate nursing students) assigned to each nursing home obtained informed consent from residents or their legal representatives. This study was approved by the participating university’s committee for the protection of human subjects in research.
The conceptual model of PD (Brown & Sears, 1993) guided the selection of variables to be compared between the MDS and medical and nursing home records. In the model, factors associated with PD are grouped into three main categories:
- Tissue tolerance.
- Perineal environment.
- Toileting ability.
Indicators of tissue tolerance include factors such as age, health problems, problems with nutrition, perfusion, oxygenation, and fever. Factors in the perineal environment include frequency and kind of incontinence, mechanical chafing, skin irritants, problems of skin hydration, and fecal enzymes. Factors of toileting ability include problems of mobility, neurosensory deficits, and cognitive awareness.
The model was originally developed for a study of PD in hospital patients, but Bliss et al. (2006) saw a potential use for guiding analysis of PD in nursing home residents. They added the use of restraints as an indicator of toileting ability because of its high relevance to nursing home residents and operationalized its components using the MDS after the analysis described in this study was completed. Items found both on the model and in the chart were identified.
A data collection form based on Brown and Sears’ (1993) model was developed by the investigators to collect the needed information from the medical or nursing home records. The RA recorded data from more than 20 medical and nursing home forms reported during the time that coincided with completion of corresponding variables on the full MDS; examples included the nursing care plan, medical orders, treatment plans, medication administration records, nutrition assessment forms, special skin/wound assessment forms, bedside forms such as intake and output and dietary records, and progress notes. The RA entered data collected from the medical and nursing home records and the MDS into SPSS, version 13, which was also used for analysis.
To ensure reliability of the RAs’ data collection, the principal investigator (D.Z.B.) randomly selected one resident from each nursing home and independently recorded the same kind of data from the resident’s medical and nursing records as did the data collectors.
A percentage agreement was calculated between data from individual items on the MDS and nursing home records. A kappa statistic was used to determine overall agreement beyond chance between data obtained from all of the items on the MDS and all of the items collected from the nursing home records of the study participants and between the data collected by the principal investigator and each RA. Interpretations of kappa statistics show that values greater than 0.7 are considered very good and those 0.9 and greater are considered excellent (Feinstein & Cicchetti, 1990).
The sample included 43 residents (23 from one nursing home and 20 from the other), was 93% Caucasian and 88% female, and consisted of residents from all three floors of the nursing homes. Table 1 provides the main characteristics of the sample according to the three factors in the conceptual model of PD (Brown & Sears, 1993). In the tissue tolerance category, the sample included age (mean age = 86, SD = 8 years), the presence of comorbid health problems (e.g., Alzheimer’s disease, stroke), and nutrition problems. In the perineal environment category, both urinary and fecal incontinence was most common and present in approximately two thirds of the residents. A high percentage of residents wore absorbent products. In the toileting ability category, few residents (2%) were bedfast, and nearly half of the sample was totally dependent in toileting and hygiene care. Restraints were used with 9% of the residents.
Table 1: Characteristics of Nursing Home Residents According to MDS Factors Associated with Perineal Dermatitis (N = 43)
Accurate documentation in residents’ charts and on the MDS are important for planning resident care and supporting nursing research.
Table 2 shows the percentage of agreement between the individual items on the MDS and the medical and nursing home records of both nursing homes’ residents. Variables are arranged according to the three main categories in the conceptual model of factors associated with PD (Brown & Sears, 1993). Two items had 100% agreement, and all items except two (hypertension [54%] and eating <25% of food at most meals [61%]) showed an agreement of at least 70%. The highest percentage of agreement between the MDS and the charts (≥95%) was found for the items of age, fever, weight gain or loss, parenteral feedings, oxygenation (oxygen therapy), fecal incontinence, and paraplegia or quadriplegia.
Table 2: Agreement Between Data in MDS and Medical/nursing Home Records on Factors Related to Perineal Dermatitis
The kappa statistic (standard error) for overall agreement beyond chance of all of the items compared was 0.72 (0.03) (p < 0.001, 95% confidence intervals = 0.66 to 0.78). These results indicate very good agreement. The kappa statistic for the agreement between the principal investigator and the RA was 1 (p = 0.000007) in one nursing home and 0.72 (p = 0.00006) in the other nursing home. These results show excellent agreement.
Previous studies have demonstrated the importance of validating MDS items for use in research by comparing the MDS items with nursing home records. The findings of this study show a very good agreement between documentation in the nursing home charts and the MDS items selected and, therefore, support collecting data related to PD using the MDS. These findings are in agreement with comparisons of nursing home chart and MDS data reported by others (Hill-Westmoreland & Gruber-Baldini, 2005; Schnelle et al., 2003). Resident survey data or observer-collected data may yield different results. The conceptual model of PD was useful in guiding data collection. Not unlike other studies that focused on a small number of nursing homes, the generalizability of this study was limited by the small sample.
Clinical implications of the findings indicate that accurate documentation in residents’ charts and on the MDS are important for planning resident care and supporting nursing research. Nurse documentation on the MDS may be used for research studies and thus to improve resident outcomes and care processes. The numerous and interdisciplinary forms on which resident data are documented support the need for an organized approach and quality check when completing the MDS.
PD is one of the main complications of incontinence and increases the cost of health care. The MDS contains data about factors associated with PD identified in a published conceptual model of PD. The purpose of this study was to determine the validity of MDS data related to PD risk factors by comparing them with data in nursing home chart records. Findings indicated that MDS items defining factors associated with PD were valid and supported use of the MDS in further investigation of a significant, costly, and understudied health problem of nursing home residents.
- Bliss, D, Savik, K, Harms, S, Fan, Q & Wyman, J2006. Prevalence and correlates of perineal dermatitis in nursing home residents. Nursing Research, 55, 243–251. doi:10.1097/00006199-200607000-00004 [CrossRef] doi:10.1097/00006199-200607000-00004 [CrossRef] doi:10.1097/00006199-200607000-00004 [CrossRef]
- Brown, DS & Sears, M1993. Perineal dermatitis: A conceptual framework. Ostomy and Wound Management, 397, 20–22, 24–25.
- Casten, R, Lawton, MP, Parmelee, PA & Kleban, MH1998. Psychometric characteristics of the minimum data set I. Confirmatory factor analysis. Journal of the American Geriatrics Society, 46, 726–735.
- Centers for Medicare & Medicaid Services2006. How data are collected. Retrieved January12, 2008, from http://www.cms.hhs.gov/QualityInitiativesGenInfo/05_HowDataAreCollected.asp
- Cohen-Mansfield, J2004. The adequacy of the minimum data set assessment of pain in cognitively impaired nursing home residents. Journal of Pain and Symptom Management, 27, 343–351. doi:10.1016/j.jpainsymman.2004.01.001 [CrossRef] doi:10.1016/j.jpainsymman.2004.01.001 [CrossRef] doi:10.1016/j.jpainsymman.2004.01.001 [CrossRef]
- Cohen-Mansfield, J, Taylor, L, McConnell, D & Horton, D1998. Estimating the cognitive ability of nursing home residents from the minimum data set. Outcomes Management for Nursing Practice, 3, 43–46.
- Crogan, NL & Corbett, CF2002. Predicting malnutrition in nursing home residents using the minimum data set. Geriatric Nursing, 23, 224–226. doi:10.1067/mgn.2002.126972 [CrossRef] doi:10.1067/mgn.2002.126972 [CrossRef] doi:10.1067/mgn.2002.126972 [CrossRef]
- Feinstein, AR & Cicchetti, DV1990. High agreement but low kappa: I. The problems of two paradoxes. Journal of Clinical Epidemiology, 43, 543–549. doi:10.1016/0895-4356(90)90158-L [CrossRef] doi:10.1016/0895-4356(90)90158-L [CrossRef] doi:10.1016/0895-4356(90)90158-L [CrossRef]
- Gruber-Baldini, AL, Zimmerman, SI, Mortimore, E & Magaziner, J2000. The validity of the minimum data set in measuring the cognitive impairment of persons admitted to nursing homes. Journal of the American Geriatrics Society, 48, 1601–1606.
- Hill-Westmoreland, EE & Gruber-Baldini, AL2005. Falls documentation in nursing homes: Agreement between the minimum data set and chart abstractions of medical and nursing documentation. Journal of the American Geriatrics Society, 53, 268–273. doi:10.1111/j.1532-5415.2005.53113.x [CrossRef] doi:10.1111/j.1532-5415.2005.53113.x [CrossRef] doi:10.1111/j.1532-5415.2005.53113.x [CrossRef]
- Lawton, MP, Casten, R, Parmelee, PA, Van Haitsma, K & Kleban, MH1998. Psychometric characteristics of the minimum data set II: Validity. Journal of the American Geriatrics Society, 46, 736–744.
- Meeks, S2004. Further evaluation of the MDS depression scale versus the geriatric depression scale among nursing home residents. Journal of Mental Health and Aging, 10, 325–335
- Morris, JN, Jones, RN, Fries, BE & Hirdes, JP2004. Convergent validity of minimum data set-based performance quality indicators in postacute care settings. American Journal of Medical Quality, 19, 242–247. doi:10.1177/106286060401900603 [CrossRef] doi:10.1177/106286060401900603 [CrossRef] doi:10.1177/106286060401900603 [CrossRef]
- Polit, D & Beck, C2004. Nursing research: Principles and methods (7th ed.). Philadelphia: Lippincott, Williams and Wilkins.
- Savik, K, Fan, Q, Bliss, D & Harms, S2005. Preparing a large data set for analysis: Using the minimum data set to study perineal dermatitis. Journal of Advanced Nursing, 52, 339–409. doi:10.1111/j.1365-2648.2005.03604.x [CrossRef] doi:10.1111/j.1365-2648.2005.03604.x [CrossRef] doi:10.1111/j.1365-2648.2005.03604.x [CrossRef]
- Schnelle, JF, Cadogan, MP, Yoshii, J, Al-Samarrai, NR, Osterweil, D & Bates-Jensen, BM et al. 2003. The minimum data set urinary incontinence quality indicators: Do they reflect differences in care processes related to incontinence?Medical Care, 41, 909–922. doi:10.1097/00005650-200308000-00005 [CrossRef] doi:10.1097/00005650-200308000-00005 [CrossRef] doi:10.1097/00005650-200308000-00005 [CrossRef]
- Wilson, L, Brown, JS, Shin, GP, Luc, KO & Subak, LL2001. Annual direct cost of urinary incontinence. Obstetrics and Gynecology, 98, 398–406. doi:10.1016/S0029-7844(01)01464-8 [CrossRef] doi:10.1016/S0029-7844(01)01464-8 [CrossRef] doi:10.1016/S0029-7844(01)01464-8 [CrossRef]
Characteristics of Nursing Home Residents According to MDS Factors Associated with Perineal Dermatitis (N = 43)
|Characteristic||Number of Residents (%)||MDS Code|
|I. Tissue tolerance factors|
|Age (mean age = 86, SD = 8 years)||AA3|
| Alzheimer’s disease||19 (44)||I1q|
| Arthritis||10 (23)||I1l|
| Stroke||17 (40)||I1t|
| Diabetes||3 (7)||I1a|
| Hypertension||22 (51)||I1h|
| Macular degeneration||4 (9)||I1mm|
| Osteoporosis||15 (35)||I1o|
| Parkinson’s disease||3 (7)||I1y|
| Cancer||4 (9)||I1pp|
| Peripheral vascular disease||2 (5)||I1j|
| Antibiotic medications||43 (100)||O4|
| Antipsychotic/hypnotic agents||8 (19)||O4a|
| Diuretic agents||13 (30)||O4e|
| Antidepressant agents||23 (53)||O4c|
| Fever||0 (0)||J1h|
| Weight gain or loss >3 pounds in 7 days or 10% in past 180 days||0 (0)||J1a|
| Edema||6 (14)||J1g|
| 5% weight gain or loss in past 30 days||10 (23)||K3b|
| Eats <25% of food at most meals||19 (44)||K4c|
| Dietary supplement||7 (16)||K5f|
| Planned weight change program||8 (19)||K5h|
| Mechanically altered diet||12 (28)||K5c|
| Parenteral/enteral feedings||7 (16)||K6a, b|
| Oxygen therapy||1 (2)||P1ag|
|II. Perineal environment factors|
| Urinary only||12 (28)||H1b|
| Fecal only||2 (5)||H1a|
| Pads or briefs used||41 (95)||H3g|
|III. Toileting ability factors|
|Sliding board use||6 (14)||G6e|
|Manually lifted||9 (21)||G6c|
|No locomotion on unit||1 (2)||G1e|
|Cane, walker, or crutch use||13 (30)||G5a|
|Inability to walk in room||23 (53)||H1|
|Wheelchair use||28 (65)||G5b, c, d|
|Total dependence with eating||9 (21)||G1h|
|Total dependence with bathing||23 (53)||G2|
|Total dependence with toileting||18 (42)||G1i|
|Total dependence with hygiene care||21 (49)||G1j|
|Restraints used||4 (9)||P4|
| Hemiplegia||5 (12)||I1v|
| Quadriplegia||0 (0)||I1z|
| Level of consciousness (alert)||39 (91)||B1, B4|
| Ability to make self understood||31 (72)||C4|
| Memory intact||10 (23)||B2|
Agreement Between Data in MDS and Medical/nursing Home Records on Factors Related to Perineal Dermatitis
|I. Tissue Tolerance Factors|
| Alzheimer’s disease||70|
| Macular degeneration||79|
| Parkinson’s disease||91|
| Peripheral vascular disease||93|
| Antibiotic medications||93|
| Antipsychotic/hypnotic agents||74|
| Diuretic agents||93|
| Antidepressant agents||88|
| Weight gain or loss >3 lbs in 7 days or 10% in past 180 days||98|
| 5% weight gain or loss in past 30 days||83|
| Eats <25% of food at most meals||61|
| Dietary supplement||72|
| Planned weight change program||81|
| Mechanically altered diet||86|
| Parenteral feedings||98|
| Enteral feedings||91|
| Oxygen therapy||98|
|II. Perineal environment factors|
| Urinary only||86|
| Fecal only||98|
| Pads or briefs used||93|
|III. Toileting ability factors|
|Sliding board use||84|
|Cane, walker, or crutch use||77|
|Inability to walk in room||93|
|Total dependence with eating||74|
|Total dependence with bathing||91|
|Total dependence with toileting||86|
|Total dependence with hygiene care||91|
| Level of consciousness (alert)||86|
| Ability to make self understood||77|
| Memory intact||76|