Web Table 1: Characteristics of Included Studies, Full Details of Intervention Studies

Systematic review: reducing dehydration riskResubmission 1, 26/10/2014 Submission 27/07/2014

Web Table 1: Characteristics of included studies, full details of intervention studies

Author / Ethical review / Study design, aims and methods / Participants: selection & baseline characteristics / Intervention, Control and Duration / Description of outcome measure & definition of low fluid intake or dehydration used by authors
*Allen,1,2
2011
UK / Yes / Design:
RCT, unblinded.
Web-based random numbers table used for allocation following consent
Aim(s) examined in this review:
To analyse the influence of serving method on consumption of ONS drinks in older adults.
Recruitment methods:potential participants identified by facility staff
Setting:
Nursing home. Size N/R / Baseline characteristics, Groups 1 (n=8) & 2(n=16) combined:
Age, years, mean (SD): 86.3 (8.9), range: 66-97
MMSE, mean (SD), range: 11.3 (9.7), 0-30
Gender, functional ability: not reported
Inclusion criteria:
Consultee consent required due to cognitive impairment
>65 years
Exclusion criteria:
Unstable DM
Investigations for DM
Dysphagia requiring thickened fluids
Lactose intolerance
Coeliac disease
Allergy to ONS
Delirium
Fluid restriction <1l/day / Group 1:
Straw inserted into ONS bottle.
Group 2:
ONS decanted into a glass/beaker.
Duration:
x3 per day, alternate days one week. / Outcome measure(s):
Proportion of ONS consumed, per number of drinks served.
Method of assessment:
Amount of ONS consumed estimated as a proportion of amount served (0.1, 0.25, 0.5, 0.75, 1.0). Method of estimation not described.
Outcome assessors: research staff
Definition of low fluid intake: N/R
*Allison,3
2005
US / N/R / Design:
Before-afterPretest-posttest, non-randomised, uncontrolled, unblinded
Aim(s) examined in this review:
To assess effectiveness of nursing practices to increase hydration in residents with TBR >550ohms.
Recruitment method, main study:
1225 residents selected by reviewing medical records
Recruitment method, sub-study included in this review:N/R
Setting:
26 long-term care facilities (17 urban, 9 rural), Texas, US. Size of facilities N/R / Baseline characteristics, main study, n=1225:
Age, years, mean:76 (SD N/R), range: 20-104; 89.4% ≥60 years.
Males: 471(38.4%)
Body weight, kg, mean: 71.89 (SD N/R)
Cognitive ability: N/R
Functional ability: N/R
Baseline characteristics, sub-study included in this review:
n=281 recruited to study and n=198 completed with outcomes reported. Characteristics of sub-study group N/R separately from the main study
Inclusion criteria, main study:
Residents having a profile of non-invasive measures for clinical management purposes.
Inclusion criteria, sub-study:
TBR readings >550ohms
Consent. / Senior facility staff evaluated participants and intervened with the appropriate care (not described)if required, to improve hydration
Duration:
3-4 months / Outcome measure(s):
Change in BIA reading of TBR.
Method of assessment:
Quantum II BioimpedenceAnlayzer. Software:Cyprus Body Composition Software system - RJL Systems
Equation: TBW/(height2/TBR) x height/Resistance
Outcome assessors: N/R
Definition of low TBR:
Hypovolemic risk and risk of dehydration defined as TBR >550ohms (average TBR: 504.81 ohms).
*Cleary,4
2008.
Canada / N/R / Design:
Before-afterPretest-posttest, non-randomised, uncontrolled, unblinded
Aim(s) examined in this review:
To evaluate effects of routine mealtime seating plan on fluid intake in residents with dementia and at risk of nutritional decline.
Recruitment method:N/R
Setting:
Long-term care facility. Size N/R / Baseline characteristics, n=4:
(n=1 RIP during the study; study reported on n=3)
Age, years, mean: 92.7 (SD N/R); range, 88-97
Sex: N/R
Diet level: pureed diet & nectar-thick fluids, n=2; regular texture & thin liquids, n=1.
MMSE, mean: 11.0(SD N/R), range: 5-14
Functional ability: N/R
Inclusion criteria:
Diagnosis of dementia
Adequate vision and hearing for normal conversation
Resident for ≥3 months
Risk of nutritional decline
Meals taken in the dining room
Able to self-feed>25% of meal for at least 2 days per week
Exclusion criteria:
Medications to stimulate diet
Depression / Regular seatingplan instituted for this study at lunchtimes.
Duration:N/R, but 12 meals observed over non-consecutive days / Outcome measure(s):
Change in percentage of fluids consumed at lunch as a proportion of amount served (amount served N/R)
Outcome assessors: N/R
Method of assessment:N/R
Definition of low fluid intake: N/R
Dunne,5
2004
US / N/R / Design:
Before-afterPretest-posttest, non-randomised, uncontrolled, unblinded
Aim(s) examined in this review:
To investigate whether using high and low contrast tableware increased fluidintake in males with advanced AD
Recruitment method:N/R
Setting:
Long-term care unit. Size N/R / Baseline characteristics, study 1, n=9:
Age, years, mean:82.7 (SD N/R), range: 72-89
MMSE, mean: 2.9(SD N/R), range 0-8
Functional ability: N/R
Visual abilities (acuity, colour discrimination, contrast sensitivity): assessment attempted, but no participant was able to complete the testing.
Baseline characteristics, study 2, n=9 (included 5 participants from Study 1; n=3 RIP during the study).
Participant characteristics N/R fully, but the 4 new men were described as being matched for age, education level and MMSE with the 4 participants they replaced.
Age, mean: 83.1 (range and SD N/R)
MMSE, mean: 3.2 (range and SD N/R)
Inclusion criteria, both studies:
Advanced AD
Able to eat independently
Exclusion criteria, both studies:
Serious medical conditions
Uncontrolled seizures
Hypertension
Diabetes mellitus
Neurological disease. / Study 1:white tableware (control), high-contrast red tableware, white
Duration: 30 days (10 days each)
Study 2 (1 year later):white tableware, high-contrast blue (n=9), white, low-contrast red (n=7), white, low-contrast blue (n=6), white
Duration:70 days (10 days each) / Outcome measure(s):
Change in mean percentage of fluid intake.
Method of assessment:
Food and fluid intake recorded every day for each participant at lunch
Amount consumed expressed as a percentage of amount served.Amount served was weighed in ounces.
Outcome assessors: N/R
Definition of low fluid intake: N/R
Fries,6
1997
US / N/R / Design:
Before-afterPretest-posttest, uncontrolled, unblinded
Aim(s) examined in this review:
To evaluate effect of implementing RAI-MDS on prevalence of dehydration at baseline and numbers of participantsdeclining or improving during 6-month follow-up period
Pre-RAI, recruitment method, nursing homes:
Purposive sampling to select facilities representing geographical location, size (>25 beds), urban/rural setting; high/low staffing levels; Medicaid reimbursement level; re-imbursement methodology (case mix/non-case mix).
Post-RAI, recruitment method, nursing homes:
Same nursing homes, but 14 lost to follow-up (10 declined, 3 closed, 1 ‘ineligible’, reason N/R).
Pre-RAI, recruitment method, residents:
8-16 residents (dependant on size of facility) stratified by length of stay (25% 30-60 days, 75% >60 days)selected from resident census at each facility on the first day of data collection by research staff.
Post-RAI, recruitment method, residents:
8-16 residents randomly selected from resident census at each facility on the first day of data collection by research staff (not stratified by length of stay and details of randomisation N/R).
Setting:
268 nursing homes in 10 US states. / Baseline characteristics, pre/post intervention:
Pre-intervention (1990): n=2128
Post intervention (1993): n=2088 (new cohort).
Baseline characteristics were combined for both the pre/post intervention groups and reported as raw frequency and weighted percentages of the total population they represent(n=121,337):
Age, n (%):
<65 years n=308 (6.9)
65-74 years n=513 (11.5)
75-84 years n=1384 (33.2)
>85 years n=2029 (48.4)
Males, n (%): 1026 (23.3)
CPS, n (%):
Intact:1296(30.9)
Moderate: 1507(34.6)
Severe: 1407(34.5)
ADL, n(%):
Independent: 770(19.0)
Limited: 1279 (29.6)
Dependent: 2150 (51.4) / Implementation of RAI-MDS, commenced in USA Autumn 1990. / Outcome measure(s):
Change in baseline dehydration prevalence.
Change in number of residents acquiring dehydration or improving during 6 months follow-up.
Method of assessment & definition of dehydration:
Dehydration present/absent as defined by the RAI-MDS, ≥2 criteria present from the following:
Fluid intake <1.5l/day
Clinical signs of dehydration
Fluid loss > fluid intake.
Methods used to assess these N/R.
Outcome assessors: research staff
Holzapfel,7
1997
US / N/R / Design:
RCTc, unblinded
Details of randomisation N/R
Aim(s) examined in this review:
To determine whether the position adopted bythe feedingassistant influences food and fluid intake in residents requiring complete feeding assistance.
Recruitment method:
N/R, but participants randomly assigned to one of three groups.
Setting:
240-bed nursing home unit in a 1000-bed medical centre. / Baseline characteristics, reported for the 3 groups combined, n=39:
Age, years, mean (range): 75 (40-99). 95% aged ≥60 years.
Males: 3 (8%).
Dementia diagnosis: 22 (56.4%).
Functional ability: N/R
Inclusion criteria:
Complete feeding assistance required for each meal
Exclusion criteria:
Enteral feeding.
Feeding assistance provided by non-designated feeders. / Intervention groups: Feeding assistants sat for 2 weeks, then stood for 2 weeks, then crossed over.
Control:Feeding assistants chose positions (positions chosen N/R)
Duration: Lunch, Monday – Friday, 4 weeks / Outcome measure(s):
Differences in mean fluid intake between the three groups on days 1,5,10,15,20 of the study. Results presented as t statistic and p value.
Method of assessment:
Food and fluid consumed,at mid-day meal only, was recorded by the feeder using four pre-determined categories of percentages (0-25%, 26-50%, 51-75%, 76-100%). Not known ifthis was measured or estimated.
Fluid defined as being able to be consumed through a straw.
Outcome assessors: facility staff delivering intervention.
Definition of low fluid intake: N/R
*Kenkman,8
2010
UK / Yes / Design:
Cluster CCT, non-randomised, unblinded
Aim(s) examined in this review:
To assess the health, well-being & nutritional status of older people living in residential care.
Recruitment method, homes:
3 residential homes already selected for development by Norfolk County Council (NCC). These were matched by resident population size with 3 other homes owned by NCC.
Recruitment method, resident:
n=213 approached
n=105 case note review (63 at year 1)
n=87 interviews (56 completed 1 year follow-up interviews, and this sub-study reported here)
n=42 RIP/moved away/declined / Baseline characteristics for intervention group completing 1 year follow-up, n=30 (sub-study):
MMSE, mean(SD): 19(5.6)
Residents attempting chair-stands: 6 (20%)
Age sex N/R for this group separately, but characteristics from main study (n=57):
Age, mean(SD): 86.1(6.7).
Males: 19 (33%)
Baseline characteristics for control group, completing 1 year follow-up, n=26 (sub-study):
MMSE, mean(SD): 17(6.2). p=0.21†
Residents attempting chair-stands: 4 (15.4%). p=0.65†
Age sex N/R for this group separately, but characteristics from main study (n=48):
Age, mean(SD): 87.7(6.8), p=0.23
Males: 12 (25%), p=0.35
Inclusion criteria, residents:
Consent
Exclusion criteria, residents:
Short stay / Intervention: Improved comfort for mealtimes, including restaurant atmosphere & extended mealtimes, increased choice of foods, Improved social experience, visitors encouraged to eat at care home, increased encouragement to eat, increased availability of drinks and snacks to residents and visitors
Control:‘Usual care’ (not described).
Duration:12 months / Outcome measure(s) ‡:
Change in number of residents with dehydration.
Method of assessment & definition of dehydration:
Presence of either:
Dry, furrowed tongue
Dry mucous membrane
Sunken eyes
Methods used to assess these N/R
Outcome assessors: research staff
Lin,9
2013
Taiwan / Yes / Design:
Cluster CCT, non-randomised, unblinded.
Aim(s) examined in this review:
Pilot study to assess whether increased fluid intake reduces asymptomatic bacteria in incontinent nursing home residents over 6 weeks.
Recruitment method, nursing homes:
Voluntary group assignment (homes chose whether they were intervention or control)
Recruitment method, residents:
n=86 eligible(identified by nursing home staff)
n=6 withdrew (2 hospitalised, 4 RIP)
n=74 completed
Setting:
6 nursing homes (30-120 beds). / Baseline characteristics, intervention group, n=44:
Age, years, mean (SD): 75.5 (12.7)
Males:14 (31.8%)
SPMSQ, mean(SD): 5.4(3.6)
Barthel Index, mean(SD): 28.6 (24.4)
MNA, mean(SD): 19.3(3.0)
Occasional incontinence:9(20.5%)
Frequent incontinence:13(29.5%)
History of UTI: 19(43.2%)
Baseline characteristics, control group, n=30:
Age, years, mean (SD): 74.7 (10.8), p=0.769†
Males: 15 (50%), p=0.116†
SPMSQ, mean(SD): 6.6(3.2), p=0.13†
Barthel Index, mean(SD): 32.2 (24.8), p=0.545†
MNA, mean(SD): 18.4(4.4), p=0.332†
Occasional incontinence: 8 (26.7%)p=0.53
Frequent incontinence: 8 (26.7%), p=0.79†
History of UTI: 12(40%), p=0.785†
Inclusion criteria:
Urinary incontinence
Exclusion criteria:
Indwelling urinary catheter
Severe mental impairment
Current urological disease
Current UTI
Current antibiotics
Inability to produce 2 urine specimens / Intervention group:advice to increase fluid intake to >1500mls/day (description of advice provide N/R), unrestricted choice of drinks.
Control group:unrestricted drinks, residents could choose type and amount.
Duration:6 weeks / Outcome measure(s) ‡:
Change in mean fluid intake.
Methods of assessment:
Fluid input/output charts completed by facility staff, but methods of measuring fluids N/R.
Outcome assessors: facility staff
Definition of low fluid intake:
Adequate fluid intake not defined, but assumed that insufficient fluid intake ≤1500mls/day, as the aim was to increase fluids to ≥1500mls/day.
*McCormick10,11
2006
Ireland / Stated as not required / Design:
RCTc, un-blinded
Aim(s) examined in this review:
In residents with dysphagia, fluid intake was compared between residents receiving drinks, thickened at the bedside by staff compared with commercially pre-prepared thickened drinks.
Recruitment method:N/R
Setting:
Long-term care unit within a 321-bedfaciltity which included acute and rehabilitation units. / Baseline characteristics, n=11:
Age, years, mean (range): 76 (51-109)
Males: 3(27%)
MMSE administered, results N/R
Barthel index, mean: 0.4 (SD N/R)
Inclusion criteria:
Dysphagia, at risk of aspiration / Weeks 1-6:Group A receivedcommercially-prepared pre-thickened drinks, Group B received drinks thickened at bedside
Weeks 7-12: Group A: drinks thickened at bedside, Group B: commercially-prepared pre-thickened drinks
Duration: 12 weeks / Outcome measure(s) ‡:
Difference in amount of thickened fluids consumed
Method of assessment:
Daily assessment of total fluid intake using graduated cups
Outcome assessors: facility staff
Definition of low fluid intake: N/R
*Mentes,12–14
2000
USA / Yes / Design:
Cluster RCT, unblinded
Aim(s) examined in this review:
To test the effectiveness of a two month hydration management guideline in preventing or reducing acute confusion and other hydration-linked events (HLE’s).
Recruitment method, nursing homes:
Methods N/R, except that a coin toss was used to randomise homes.
Larger sites, residents selected from certain units. Smaller sites: selected from whole site.
Recruitment method, residents:
Stratifiedsampling to obtain gender balance, details of approach N/R. Participants recruited following randomisation of homes.
Setting:
4 nursing homes (size: 84-717 beds). 1 smaller community home 1 larger site in each arm of the trial. / Baseline characteristics, intervention group, n=25:
Age, years, mean (SD): 80.6 (10.3)
Males: 11 (44%)
BMI, mean(SD): 26.1(6.8)
MMSE, mean (SD):22.0 (5.6)
NEECHAM, mean (SD): 26.4 (2.8)
FIM, mean (SD): 79.4 (22.3)
Baseline characteristics, control group, n=24:
Age, years, mean (SD): 83.0 (9.2), p=0.41
Males: 11 (46%), p=0.89
BMI, mean (SD): 27.5 (5.9), p=0.15
MMSE, mean (SD): 24.6 (3.6), p=0.061
NEECHAM, mean (SD): 28.3 (1.4). p=0.003
FIM, mean (SD): 112.2 (10.9), p<0.001
Inclusion criteria:
≥65 years
Able to participate in baseline & outcome measures
Consent
Exclusion criteria:
Current acute confusion
Terminal illness
Current uncontrolled DM
GI conditions affecting transit & absorption of fluids
Severe CCF
Renal failure necessitating fluid restriction or creatinine levels >=3.5mg/dl
Serum Na <135mEq/l
Current UTI / Intervention:research staff calculated personal fluid goals, based on weight, and 75% of this goal to be achieved at mealtimes. Participants provided with increased choice and availability of drinks, ‘sip-and-go’ cups and tagging of charts and trays for ‘higher-risk’ residents.
Control group:routine care (not described).
Both groups: care staff trained in detecting and documenting acute confusion.
Duration:8 weeks, but unclear if it was 5 or 7 days a week. / Outcome measure(s) ‡:
1. Change in urine colour.
2. Change in USG
3. Change in fluid intake
4. Number of participants achieving fluid intake of >=75% of fluid goal.
5. Change in TBW.
Method of assessment:
1. Standard urine colour chart
2. USG, assessed using Chemstrip Mini Urine Analyser
Weekly urine assessments.
3. 2x 24 hour fluid intake records documented at baseline. During each week of the intervention a partial fluid intake record of drinks taken at mealtimes, medication and fluid rounds, was documented. Caffeinated and alcoholic beverages were excluded. Method of assessment N/R.
4.BIA-101 Quantum Analyser (RJL systems, Clinton Township). Equations referenced.
BIA conducted at baseline, weeks 4 and 8.
Outcome assessors: facility staff
Definition of low fluid intake and dehydration:
1. Urine colour: higher colours (7,8) indicate dehydration
2. USG, mild dehydration: USG >1.020; dehydration: ≥1.025.
3. Calculation of weight-based goal for fluid requirements: 100ml/kg for the first 10kg, 50ml/kg for the next 10kg, 15ml/kg for remaining body weight. Fluid intake 75% of calculated goal described as inadequate.
4. TBW normal values, 55-65% in lean to normal males; 50-60% in lean to normal females ('lean to normal' described as BMI <30.6).
Robinson,15 2002
US / N/R / Design:
Before-afterPretest-posttest, non-randomised, uncontrolled, unblinded.
Aim(s) examined in this review:
To determine the effect of a hydration programme on the level of hydration and the prevention of conditions associated with dehydration.
Recruitment method:N/R
Setting:
1 nursing home (250 beds). / Baseline characteristics, n=51:
Age, years, mean:83.5(SD N/R), range: 66-97
Males: 8 (15.7%)
Number of participants with TBW lower than the standard, n (%): 24 (47)
Dependency level n(%):
Complete: 9 (17.6)
Holds glass: 25 (49)
Independent:17(33.3)
Thickened drinks: 7(13.7)
Cognition level: N/R
Inclusion criteria:
Consent / Intervention (7 days/week, 5 weeks): Goal: to drink 8oz more fluids twice a day at mid-morning & afternoon, hydration assistant for fluid administration, colourful beverage cart, jugs & glasses, choice of 4 drinks, special requests honoured
Duration:9 weeks (2 weeks baseline, 5 weeks intervention, 2 weeks follow-up) / Outcome measure(s)3:
Number of participants who always drank the extra 16 floz/day.
Change in TBW
Method of assessment:
Fluid intake monitored mid-morning & afternoon only.
Use of BIA to assess TBW, methods not described, and information regarding type of machine, and equations used N/R
Outcome assessors: research staff
Definition of low fluid intake and dehydration:
Inadequate fluid intake defined as ≤1500mls/day
Low TBW described as being less than ’the standard’; ‘the standard’ described as derived from the National Health & Nutrition Examination Survey.
*Schnelle,16 2010
US / Yes / Design:
RCT, unblinded
Random numbers table used
Aim(s) examined in this review:
To evaluate the effects of a multi-component intervention on FI and UI incontinence in incontinent residents.
Recruitment method, residents:
n=495 eligible(identified by care staff)
n=153 consented
n=143 baseline assessments
n=125 randomised
n=112 completed trial
Setting:
6 nursing homes. Size N/R. / Baseline characteristics intervention group, n=58:
Age, years, mean (SD): 85.8 (9.4)
Males: 16%
MMSE, mean (SD): 12.9(8.4)
Sit-to-stands, number performed, mean (SD): 3.6(3.1)
Walk or wheel distance, meters, mean(SD): 58.7(50.3)
Baseline characteristics control group, n=54:
Age, years, mean (SD): 86.1 (10.5), p=0.87
Males:19%, p=0.68
MMSE, mean (SD): 9.6(8.4), p=0.04