Inadequate water intake frequently is cited in gerontological nursing textbooks as contributing to a number of health problems of elderly individuals including acute confusion, incontinence, constipation, and adverse drug responses. However, research which documents the water intake of elderly individuals and the association of water intake with environmental and subject characteristics is limited. For the purposes of this study, water is defined as the water content of food and fluid ingested, while fluid intake is computed from the liquids ingested.
Until the water intake of nursing home residents is documented and at-risk factors for inadequate water intake are identified, nursing interventions directed at prevention of the problem are hampered. Therefore, the purposes of this study were to:
* Determine the water intake of nursing home residents.
* Identify variables associated with adequacy of water intake.
Water, a critical nutrient for life, is necessary in the maintenance of individuals' homeostasis. Water is obtained from fluids, foods, and the oxidation of foods ingested. Water intake should be equivalent to the quantity of water secreted plus insensible losses. Daily water needs have been defined as 2,500 mL of water daily for the normal circulation of healthy adults; however, this standard does not consider body size or age. Considering the water intake from fluids only, 1,500 mL daily has been recognized clinically as a minimum (Reedy, 1988).
Norton, McLaren, and ExtonSmith (1962) documented the lack of fluid intake among 18 geriatric longstay patients in an English hospital. They found that all but one of the 18 patients observed had a fluid intake of less than 1,500 mL in 24 hours. More recently, several studies have documented the fluid intake of elderly individuals. Hart and Adamek (1984) reported an average fluid intake of 1,600 mL per day for a group of nursing home residents in a study to prevent urinary catheter blockage. The nursing intervention consisted of increasing fluid intake to between 2,000 and 3,000 mL per day. Even with the focus on increasing fluid intake, the minimum of 2,000 mL per day was not achieved. Adams (1988) compared the fluid intake of 30 institutionalized elderly individuals and 30 community-dwelling elderly individuals. The mean daily fluid intake of the institutionalized subjects was 1,507 mL per day, compared with 2,115 mL per day for the community-dwelling group. None of these studies identified factors associated with fluid intake.
Two studies focused on water intake, rather than fluid intake (Chidester & Spangler, 1997; Gaspar, 1988). Water intake, defined as the water ingested from food and fluid, was found to be inadequate among a group of 67 rural nursing home residents (Gaspar, 1988). The mean intake of water from food and fluid was 1,892.5 mL per day, ranging from 832.5 mL per day to 2,862.5 mL per day. Only 3 (4.5%) of 67 subjects met 100% of their water requirement, based on the standard of 1,600 mL/mp 2 body surface area (BSA). Variables identified as associated with inadequate water intake were categorized as:
* Subject characteristics.
* Ingestion behaviors.
Subjects who had inadequate water intake were older than age 85 and had a longer length of stay in the nursing home. Eight or fewer ingestions in a 24-hour period was an ingestion behavior found to be associated with inadequate water intake. Two functional ability variables - needing some assistance in feeding and semidependent functional ability - were found to be associated with inadequate water intake. These variables indicate that a resident who is not completely dependent or independent in self-care is at higher risk for inadequate water intake. Two communication variables - intact speech and hearing - were associated with intake. The association reflected that those with intact speech and hearing were more likely to have a lower water intake.
Chidester and Spangler (1997) reported water intake of 40 nursing home residents. These investigators found that the mean water intake of subjects studied was 1,632 mL per day, ranging from 871 mL per day to 3,558 mL per day. Three different standards for determining water intake adequacy were compared. The standards were:
* 30 mL/kg for weight.
* 1 mL/kcal energy consumed.
* 100 mL/kg for first 10 kg, 50 mL/kg for next 10 kg, and 15 mL for remaining kg.
The majority of subjects were determined to have an inadequate water intake using all three standards. The third standard, which adjusts for body weight, was indicated as the best standard because it provided at least 1,500 mL for those with low body weights and did not overpredict for obese individuals.
In contrast to the findings of Gaspar (1988), Chidester and Spangler (1997) found no significant associations between water intake and age, cognitive skills, physical locomotion, ability to understand, and ability to feed self. The difference in data collected and in approaches to analysis between the Chidester and Spangler (1997) study and the Gaspar (1988) study needs to be considered when comparing results of the two studies. Thus, the purposes of this study were to:
* Explore the adequacy of water intake among nursing home residents.
* Identify variables associated with the adequacy of water intake.
This study expands the database of the previously mentioned study of Gaspar (1988) by adding 32 subjects from an urban nursing home, thus increasing the sample to 99 subjects. The method for the urban site study was a replication of the previous study of Gaspar (1988) conducted in two rural nursing homes. The expanded database enhances data analysis procedures for identification of variables associated with inadequate water intake. With the identification of these variables, interventions for prevention of this problem can be targeted to those residents at risk.
The study was conducted in one urban nursing home in Utah and two rural skilled nursing homes in South Dakota. The urban nursing home has 98 beds, and the rural nursing homes have 79 and 60 beds. All three nursing homes provided intermediate and skilled care and were Medicare-approved and Medicaid-approved institutions. All data were collected prior to the initiation of the Omnibus Reconciliation Act of 1987 (OBRA '87, Public Law No. 100-203). Approval to conduct the study was obtained from the Research Review Committee of the Frances Payne Bolton School of Nursing, Case Western Reserve University, the University of Utah Human Institutional Review Board, and the nursing homes where the study was conducted.
The convenience sample was comprised of nursing home residents that met the following selection criteria:
* Not on a fluid restriction.
* Not receiving tube feedings.
* 70 years of age or older.
The criteria allowed residents with a range of physical and cognitive abilities to participate. Family and guardians were contacted to obtain consent if the individual was unable to make a decision or if they had a legal guardian.
Data Collection Procedure
Data were collected primarily through nonparticipant observation. Two 24-hour periods of observation were conducted. Chart review and the quantitative measurement of weight, height, and urine output were obtained. Recording of the data was facilitated through the use of two investigator-developed datarecording instruments: the Individual Characteristics Sheet and the Intake/Situational Modifier Sheet. The observers also completed Norton's At-Risk For Pressure Sore Scale (Norton et al., 1962).
The Individual Characteristics Sheet was a recording sheet designed to elicit subjective reports of thirst, fear of incontinence, health, and nausea experienced and to facilitate the recording of individual characteristics, medication administered, height, weight, and sensory function. The basis for selection of these characteristics was to adequately describe the sample and ascertain the characteristics cited in the literature as potentially related to water or fluid intake.
Figure. Comparison of subjects' actual water intake and standard water intake requirement based on 1,600 mL/m2 BSA (N= 99).
The Intake/Situational Modifier Sheet was used by the observers throughout the observational period to record all food and fluid offered and ingested by subjects. These recordings were used to determine the actual amounts of food and fluid ingested. Situational modifiers also were recorded on this instrument. Situational modifiers are factors that describe each intake session and potentially can change with each ingestion session. The modifiers studied were:
* Level of assistance.
* Swallowing ability.
* Choking and coughing occurring with swallowing.
* Place of ingestion.
* Occurrence of spilling.
* Position of upper body and head during ingestion.
* Who initiated the ingestion.
* Occurrence of drooling.
* Availability of fluids throughout the observation.
These modifiers were selected based on a literature review and from a pilot observation as potential modifiers of water intake.
Norton's At-Risk for Pressure Sores Scale was used to measure physical and mental function (Norton et al., 1962). The scale has five components:
* General condition.
* Mental condition.
Newman and West (1981) concluded that the Norton's At-Risk for Pressure Sores Scale was useful in determining the general function of an individual. Quick (1988) compared the scale with the Katz Index of Activities of Daily Living Scale (Katz et al., 1963) and found high correlations between the subscales and total function.
Each component of the Norton's At-Risk for Pressure Sores Scale (Norton et al., 1962) is rated on a 4-point scale, with a total possible score of 20. A score of 14 or below places an individual at risk for pressure sores. The scale has been shown to be reliable in the identification of patients at risk for pressure sores (Goldstone & Goldstone, 1982; Goldstone & Roberts, 1980; Roberts & Goldstone, 1979). For the purposes of this study, a score of 20 was indicative of high functional level, and the lower the score, the greater the level of dependence.
The data collection procedure was similar at each institution. Staff at each institution were informed of the purposes of the study and that the purpose of the observation was to record subject behaviors not actions of the staff. Residents that met the subject criteria were approached to obtain consent. The subjects' or family members' or guardians' written consent was obtained. Confidentiality and privacy of subjects and individual nursing homes were protected by the assignment of a code number to each subject and nursing home. These code numbers were kept in a locked file cabinet. Also, subjects were informed of their right to withdraw from the study at any time and that lack of participation would not jeopardize their care at the facility in any way.
Among the residents who consented to be in the study, clusters of four to five subjects were formed so subjects could be observed readily in and out of their rooms. Having the subjects in the cluster eating in the same dining room enhanced the accuracy of meal-time ingestion.
Each 24-hour observation was completed in 8-hour or 12-hour shifts by the observers. The observer selected a location in or near each subject's room that was as nondisruptive as possible to the residents and staff, yet allowed accurate observations. The observer rotated freely among the rooms of the subjects. When subjects left the room for activities or meals, the observer would move continuously among the areas where the subjects were located. Following each observation shift, the observer completed the Norton's At-Risk for Pressure Sores Scale (Norton et al., 1962) and indicated the subject's visual, speech, and hearing ability. Because more than one observer rated these abilities and the observations were at different times of the day, an observer consensus of the ratings was obtained.
Data collection was completed by trained nurse research assistants. The training consisted of a process to ensure reliability of food and fluid measurements. A simulated nursing home meal situation with four residents was the final step. Simulated situations were conducted until the amount of food and fluid ingested as recorded by each observer and the actual amount (calculated by subtraction of amount remaining on tray from the amount offered) reached a .95 level of agreement.
Data Analysis Procedure
The coding of data progressed in two steps. The first step was the coding of food and fluid items necessary for the analysis of water content of foods. The Highland View Hospital and Case Western Reserve University Nutrient Data Base (Case Western Reserve University, 1983) was used to determine the water intake for food and fluid. The second step was the coding and entry of the individual characteristics, situational modifiers, urine output, and need conditions data. Data were analyzed using the Statistical Package for the Social Sciences (SPSS 7.5 for Windows) system language and operations. Frequencies and measures of central tendency and dispersion were computed for all variables and descriptive data. An index of adequacy of water intake was calculated based on the standard water requirement of 1,600 mL/m2 body surface area. A significant difference between the water intake of the observation days was not found; therefore, an average of the two 24-hour observation periods was calculated.
Sixty-seven residents of the rural nursing homes and 32 residents from the urban setting participated in the study for a total of 99 subjects. The mean age of the 23 men and 76 women was 85 years. Approximately 40% of subjects had an impairment in hearing, vision, or speech. Sixty-eight percent of subjects (n = 69) had been residents of the nursing home for 1 year or more. These demographic characteristics closely reflect the national nursing home population.
Fifty-one subjects were able to respond to the interview questions conducted at the end of the observation. Twenty subjects (39%) indicated they feared incontinence and thereby limited fluids. Thirteen subjects indicated that feelings of nausea were common and that they did not report this to the nursing staff. Among the 51 subjects that could respond, 32 (63%) indicated they had no desire for fluids. Perceived health was rated as "good" by 43% of subjects, "fair" by 37% of subjects, and "poor" by 20% of subjects.
The scores of the 99 subjects on the Norton At-Risk for Pressure Sore Scale ranged from 7 to 20, with a mean of 15. Forty-one subjects had a score of 14 or below, which is indicative of at-risk status.
The number of ingestion sessions for the 99 subjects ranged from 3 to 15 sessions in a 24-hour period, with a mean of 8 sessions in 24 hours. A session was considered a discrete time that ingestions occurred, such as a meal, medication time, or snack. For the one subject with only three ingestion sessions, the sessions were meal times. At no other time in 24 hours did this subject have anything to eat or drink. Difficulty swallowing was noted for 24 subjects, with 11 subjects experiencing drooling. Approximately three fourths of subjects (n - 76) spilled food or fluid during the 24-hour observation.
The mean water intake, considering water from food and fluid, was 1,968 mL per day. Water intake ranged from 597 mL per day to 2,988 mL per day. For each subject, this amount was compared to the standard water requirement of 1,600 mL/m2 BSA. Refer to the Figure for the comparison of the subjects' actual water intake and their standard recommended water intake. Only 8 of 99 subjects (8%) met or exceeded their required standard. The percent of actual intake compared to the standard requirement was calculated for each subject and is referred to as an indication of water intake adequacy. Actual intake ranged from 37% to 133% of the subjects' standard, with a mean of 76%. This indicated that half of the subjects met 76% or less of their required water intake based on body size.
The mean water intake from only fluids was 1,468 mL. Subjects had intakes from less than 500 mL to 2,470 mL of fluid per 24 hours. This finding indicates that more than half of the subjects had less than the recommended 1,500 mL of fluid intake per day.
ANALYSIS OF WATER INTAKE ADEQUACY WITH SELECT VARIABLES
Association of Factors With Inadequate Water Intake
Several factors were identified that placed individuals at risk for inadequate water intake. There was a negative correlation (r = -.24,/? < .05) between water intake adequacy and age. This result indicates that the older the nursing home resident the more likely to have an inadequate intake of water.
The number of ingestion sessions per day was positively correlated (r = .32, p < .01) with water intake adequacy. Nutrient intake also was found to be positively correlated with water intake. As the computation of water intake was based on food and fluid ingested, this correlation is not surprising.
Scores on the Norton At-Risk for Pressure Sore Scale and water intake adequacy were negatively correlated (r = -.1648, p = .05), indicating that the more dependent in function the higher the adequacy of water intake. Specific functional abilities that were either nominal-level or ordinal-level data were analyzed for association with water intake adequacy using nonpar ametric tests. Three specific functional abilities - speech, feeding self, and drooling - were associated significantly with inadequate water intake. For purposes of analysis, subjects were categorized into low and high groups based on the mean water intake adequacy of 76%. The low group had a water intake of less than 76% of their recommended amount, and the high group had a water intake of 76% or greater of their recommended amount. Refer to the Table for results of the analysis.
Of the 52 subjects in the low intake group, 39 subjects had no impairment in speech compared to 13 subjects with a speech impairment. Among the 16 subjects who were totally dependent in eating, 12 were in the high intake group. Forty subjects needed some assistance with eating. Twenty-five of these subjects were in the low water intake adequacy group. The presence of drooling was the third ability associated with water intake adequacy. Among the 11 subjects who drooled, only 2 were in the low water intake adequacy group. These results do not reflect a totally independent nursing home resident. Rather, they create a picture of the nursing home residents at risk for inadequate water intake as those who have intact speech, need some assistance with eating, and do not drool.
The findings of this study support the previous research finding that nursing home residents have water intake less than 100% of the standard water requirement of 1,600 mL/mp 2 BSA. The mean water intake of 1,968 mL per day is greater than the mean intake of 1,632 mL per day reported by Chidester and Spangler (1997). There was a range of water intake of 2,391 mL (597 mL to 2,988 mL) among the subjects of this study. The range of water intake for the 40 subjects reported by Chidester and Spangler (1997) was 2,687 mL (871 mL to 3,558 mL). Using different standards to determine adequacy of water intake, the majority of the subjects were categorized as having inadequate water intake.
Fluid intake of the 99 subjects was compared to the standard of 1,500 mL and was found to be inadequate. The mean intake of water from fluids only was 1,468 mL per day (489 mL per day to 2,470 mL per day). Previous research of Norton, McLaren, and Exton-Smith (1962), Adams (1988), and Hart and Adamek (1984) support these findings. It is interesting to note that the mean fluid intake found in this study and that reported by Adams (1988) among 30 institutionalized elderly individuals (1,507 mL per day) were only 39 mL per day different.
The factors associated with water intake in this study were not supported in the study of Chidester and Spangler (1997) which found no association between water intake and any of the variables of the Nursing Home Minimum Data Set that were analyzed. Specifically, age, cognitive skills, physical locomotion, ability to understand, and ability to feed self were not found to be associated with water intake.
The positive association between nutrient intake with water intake was not surprising because water for the purposes of this study was from food and fluid. Among a group of community-dwelling adults, ages 20 to 80, de Castro (1992) found the strongest predictor of the amount of fluid ingested was the amount of solid food ingested.
CONCLUSIONS AND IMPLICATIONS
The basic water need is not being met among nursing home residents. Several factors were found to be associated with inadequate water intake. These factors were:
* Older age.
* Higher score on Norton At-Risk for Pressure Sore Scale.
* Few ingestion sessions per day.
* Intact speech.
* Semidependent with eating.
* Absence of drooling.
* Inadequate nutrient intake.
One factor that was associated with water intake that can be changed is the number of ingestion sessions. The number of ingestion sessions ranged from 3 sessions to 15 sessions. The one subject with only 3 sessions only ate or drank at meal times. Contributing to the number of ingestion sessions other than meals were scheduled medication times, between-meal snacks, routine offerings of fluids, and independently initiated ingestions. Scheduling more ingestions, such as offering more between-meal snacks, especially for those identified as at risk for inadequate water intake, is a necessary intervention. This especially is important because the thirst sensation declines with age (Miller, Krebs, Neal, & Mclntyre, 1982; Phillips, Bretherton, Johnston, & Gray, 1991; Phillips et al., 1984). As institutions attempt to limit medication times, the number of ingestion sessions for nursing home residents may be decreased. This needs to be considered in the scheduling of ingestion sessions.
Other factors associated with water intake adequacy can be used to identify those residents who are at risk for inadequate intake. Nursing home residents who are older than age 85, are semidependent in eating, do not drool, and have intact speech need to be assessed for adequacy of water intake. This study presents data that support dependent residents as having a higher adequacy of water intake than those subjects who were semidependent. Nursing staff have the skill and knowledge to provide care to dependent residents. Residents who drool are recognized by staff as having the potential for inadequate intake of food and fluid; therefore, staff implement interventions to ensure adequate intake for these residents. Nursing home residents who are independent ensure their needs are met. It is those residents who are struggling to maintain independence who are not receiving the level of care they need to compensate for their limitations.
Continued research is needed. First, the standard water intake for elderly individuals has not been determined. Several current standards consider extreme weights of individuals; yet, the changing composition of the aging body is not considered. The testing of the at-risk factors identified in this study also is needed. Identification of at-risk factors will assist to target interventions for prevention of inadequate water intake. A third area is the determination of when inadequate water intake or a pattern of inadequate water intake results in consequences such as confusion, constipation, incontinence, and adverse drug responses.
The development of a protocol for the assessment, prevention, and treatment of inadequate water intake is essential. The University of Iowa Gerontological Nursing Interventions Research Center (Mentes et al., 1998) recently has addressed this need through the establishment of a research-based protocol on hydration management. The purpose of the protocol is identified as helping health care providers determine adequate fluid intake and implement strategies for maintaining hydration among older adults. At-risk factors for inadequate water intake identified in this study are incorporated into this research-based protocol. Zembrzuski (1997) emphasizes that successful implementation of a hydration program in long-term care needs to involve administration, clinical staff, and inservice education. Strategies to assist long-term care facilities in initiating a hydration program are provided (Zembrzuski, 1997).
Water is a basic human need. Ensuring adequate water intake for institutionalized elderly individuals is a responsibility of nurses. Himmelstein, Jones, and Woolhandler (1983) describe a sentinel health event as:
an illness or death that should be preventable with adequate care or should at least cause those caring for the patient to ask, "Why did it happen?" (p. 466).
Inadequate water intake among nursing home residents is a precursor to the sentinel health event of dehydration and is preventable with adequate nursing care. Nurses need to be asking why this is occurring.
- Adams, F. (1988). How much do elders drink? Geriatric Nursing, 9, 218-221.
- Case Western Reserve University. (1983). Coding manual for Highland View Hospital-Case Western Reserve University nutrient database (Revision 7). Cleveland, OH: Author.
- Chidester, J.C., & Spangler, A.A. (1997). Fluid intake in the institutionalized elderly. Journal of the American Dietetic Association, 97, 23-28.
- de Castro, J.M. (1992). Age-related changes in natural spontaneous fluid ingestion and thirst in humans. Journal of Gerontology, 47, P321-P330.
- Gaspar, P.M. (1988). What determines how much patients drink? Geriatric Nursing, 9, 221-224.
- Goldstone, L., & Goldstone, J. (1982). The Norton score: An early warning of pressure sotesi Journal of Advanced Nursing, 7, 419-426.
- Goldstone, L., & Roberts, B. (1980). A preliminary discriminate function analysis of elderly orthopedic patients who will or will not contract a pressure sore. International Journal of Nursing Studies, 17, 17-23.
- Hart, M., & Adamek, C. (1984). Do increased fluids decrease urinary stone formation? Geriatric Nursing, 5, 245-248.
- Himmelstein, D.U., Jones, A.A., & Woolhandler, S. (1983). Hypernatremic dehydration in nursing home patients: An indicator of neglect. Journal of the American Geriatrics Society, 31, 466-471.
- Katz, S., Ford, A.B., Moskowitz, R. W, Jackson, B.A., Jaffee, M.W. (1963). Studies of illness in the aged. The Index of ADL: A standardized measure of biological and psychosocial function. Journal of the American Medical Association, 185, 94.
- Mentes, J., Akins, J., CuIp, K., Gaspar, P., Kundrat, M., Mobily, P., Rapp, CG., Tripp-Reimer, T, Wadle, K., & Wakefield, B. (1998). Hydration management research-based protocol. Iowa City: University of Iowa Gerontological Nursing Interventions Research Center Research Development and Dissemination Core.
- Miller, P.D., Krebs, R.A., Neal, B.J., & Mclntyre, D.O. (1982). Hypodipsia in geriatric patients. The American Journal of Medicine, 73, 354-356.
- Newman, P., & West, J. (1981). Pressure sores-2: The value of Norton score. Nursing Times, 77, 82-84.
- Norton, D., McLaren, R., & Exton-Smith, A.N. (1962). An investigation of geriatric nursing problems in hospitals. New York: Churchill Livingstone.
- Phillips, P.A., Bretherton, M., Johnston, C.I., & Gray, L. (1991). Reduced osmotic thirst in healthy elderly men. American Journal of Physiology, 261, R166-R171.
- Phillips, P.A., Rolls, B.J., Ledingham, J.G., Forsling, M.L., Morton, J.J., Crowe, M.J., & Wollner, L. (1984). Reduced thirst after water deprivation in healthy elderly men. New EngUnd Journal of Medicine, 311, 753-759.
- Quick, D.M. (1988). Functional ability of institutionalized elderly: A six month follow-up. Unpublished master's thesis, University of Utah, Salt Lake City.
- Reedy, D.F. (1988). How can you prevent dehydration? Geriatric Nursing, 9, 224-226.
- Roberts, B., & Goldstone, L. (1979). A survey of pressure sores in the over sixties on two orthopedic wards. International Journal of Nursing Studies, 16, 355-364.
- Zembrzuski, C. (1997). A three-dimensional approach to hydration of elders: Administration, clinical staff, and in-service education. Geriatric Nursing, 18, 20-26.
ANALYSIS OF WATER INTAKE ADEQUACY WITH SELECT VARIABLES