- Click here for explanation of classification scheme.

Study 1: The purpose of this study was to conduct a construct validation study that compared the index score from SCREEN II to a dietitian's rating of nutritional risk based on medical and nutritional history, dietary intake and anthropometry. An abbreviated version of SCREEN II was also developed.

Study 2: The purpose of this study was to assess the test-retest reliability of SCREEN II.

Study 3: The purpose of this study was to assess the inter-rater reliability associated with SCREEN II.

Study 1

• 55 years of age or older
• Lived in the community (Guelph area of Ontario)
• No cognitive impairment as determined by Standardized Mini Mental State Exam; scores greater than or equal to 26.

Study 2

• Seniors from the community subsample from Study 1, OR
• Seniors who lived in the community (Guelph area of Ontario).

Study 3

• Seniors who required assistance from others to live independently in the community (Guelph area of Ontario)
• No cognitive impairment.

Study 1

• Less than 55 years of age
• Did not live in the community
• Cognitive impairment.

Study 2

• Did not participate in Study 1, OR
• Did not live in the community.

Study 3

• Did not require assistance from others to live independently in the community
• Cognitive impairment.

Recruitment

Study 1:

• Participants recruited from the community in the Guelph area of Ontario via newspaper ads, posters, invitational letters and displays at various community programs and clubs
• Participants also recruited from a geriatrician's clinic. Clinic staff identified patients who were eligible and asked them to participate.

Study 2:

• 45 seniors from the community subsample from Study 1
• 100 seniors from the community using displays at a recreation center, a mall and places of worship.

Study 3:

Participants were recruited from malls, recreation and day programs in the Guelph area.

Design

Study 1:

• Geriatrician's Clinic
  • SCREEN II, weight and blood work collected at clinic visit
  • Demographic and medical questionnaire, anthropometry and 24-hour dietary recall completed in person's home (time period from clinic visit not specified)
  • Two telephone calls within a week following completion of home visit to obtain further dietary information.
• Community
  • SCREEN II, demographic and health questionnaires mailed to participants who qualified (as determined during telephone call)
  • Anthropometry, physical measures and 24-hour dietary recall completed at assessment appointment within two weeks of completion of forms
  • Two telephone calls within week following assessment appointment to collect two more days of dietary intake data.
• Development of abbreviated version of SCREEN II
  • Development used mixed methods
    • 21 dietitians with expertise in senior nutrition reviewed SCREEN II and ranked each item on its importance to the construct of nutrition risk
    • Items ranked highest were candidates for inclusion in the abbreviated version.

Study 2:

• Community sample from Study 1
  • SCREEN II, demographic and health questionnaires mailed to participants who qualified (as determined during telephone call)
  • Anthropometry, physical measures and 24-hour dietary recall completed at assessment appointment within two weeks of completion of forms; a second SCREEN II was form completed during assessment appointment.
• Additional community sample
  • First administration of SCREEN II and demographic information completed in presence of research assistant
  • Second SCREEN II mailed, completed and returned within two weeks of first administration.

Study 3:

• Telephone administration of SCREEN II and a demographic and functional assessment questionnaire
• Second telephone administration completed by second trained researcher (time from first administration not specified).

Statistical Analysis

Study 1:

• Descriptive statistics for all samples and subgroups for comparison purposes on demographics, health and functional measures
• Receiver operating characteristics (ROC) curves were created using the median of the dietitian's risk rating as the cut-point for risk
  • Area under the curve (AUC) equals consistent with scoring of the criterion (risk/no risk)
  • Higher AUC equals increased consistency
• Development of abbreviated SCREEN II: Using dietitian ranking as dependent variable, individual SCREEN II items were regressed to determine those most predictive of the risk rating.

Study 2:

• Intra-class correlation (ICC) of scores for the entire index and individual index items.

Study 3:

• Intra-class correlation (ICC) of scores for the entire index and individual index items.

Timing of Measurements

Study 1:

• Geriatrician's Clinic:
  • SCREEN II, weight and blood work collected at clinic visit
  • Demographic and medical questionnaire, anthropometry and 24-hour dietary recall completed in person's home (time period from clinic visit not specified)
  • Two telephone calls within week following completion of home visit to obtain further dietary information
• Community:
  • SCREEN II, demographic and health questionnaires mailed to participants who qualified (as determined during telephone call)
  • Anthropometry, physical measures and 24-hour dietary recall completed at assessment appointment within two weeks of completion of forms
  • Two telephone calls within week following assessment appointment to collect two more days of dietary intake data.

Study 2:

• Community sample from Study 1:
  • SCREEN II, demographic and health questionnaires mailed to participants who qualified (as determined during telephone call)
  • Anthropometry, physical measures and 24-hour dietary recall completed at assessment appointment within two weeks of completion of forms. Second SCREEN II form completed during assessment appointment.
• Additional community sample:
  • First administration of SCREEN II and demographic information completed in presence of research assistant
  • Second SCREEN II mailed, completed and returned within two weeks of first administration.

Study 3:

• Telephone administration of SCREEN II and a demographic and functional assessment questionnaire
• Second telephone administration completed by second trained researcher (time from first administration not specified).

Dependent Variables

• Study 1: SCREEN II index; development of abbreviated SCREEN II index used the dietitian ranking as the dependent variable
• Study 2: Second completion of SCREEN II index
• Study 3: Second telephone interview completion of SCREEN II index.

Independent Variables

Study 1: Dietitian's rating of nutritional risk based on medical and nutritional history, dietary intake and anthropometry (one- point scale; one = low risk, 10 = high risk).

• Medical and Nutritional History questionnaire
  • Demographics
  • List of medical problems, including vision and hearing
  • Use of medications, vitamins, minerals, herbal remedies, meal supplements and alcohol
  • Perceived health status
  • Quality of life using SF-20
  • Doctor visits and hospitalizations in the past six months
  • Dentition
  • Swallowing problems
  • Smoking
  • Gastrointestinal problems
  • Appetite, taste and smell changes
  • Specialized diets
  • Activities of daily living
  • Use of community services (e.g., Meals on Wheels, home care, nursing, etc.)
  • Income
• Anthropometry
  • Triceps and subscapular skinfolds
  • Knee height
  • Weight
  • Mid-arm, wrist and calf circumferences
• Dietary intake: 24-hour recalls (two weekdays, one weekend day)
• Laboratory data
  • Prealbumin, N=44
  • C-reactive protein, N=50
  • Hematocrit and hemoglobin, N=57
•  Development of abbreviated SCREEN II
• individual SCREEN II items.

Study 2: First completion of SCREEN II index.

Study 3: First telephone interview completion of SCREEN II index.

Initial N

Study 1:

• 61 participants from geriatrician's clinic
• 132 participants from community.

Study 2:

• 45 participants from community subsample of Study 1
• 100 participants (additional) from community.

Study 3:

97 participants: 69 for inter-rater reliability, 28 for intra-rater reliability.

Attrition (final N):

Study 1:

• 61 participants from geriatrician's clinic
• 132 participants from community.

Study 2:

• 45 participants from community subsample of Study 1
• 100 participants (additional) from community
• Nine additional participants, not specified from which group.

Study 3:

97 participants: 69 for inter-rater reliability, 28 for intra-rater reliability.

Age

Study 1:

• N=193
  • Less than 64 years, 1.6% (N=3)
  • 65 to 74 years, 48.4% (N=93)
  • 75 to 84 years, 43.8% (N=85)
  • 85+ years, 6.3% (N=12).

Study 2:

• N=149
  • Less than 64 years, 16.1% (N=24)
  • 65 to 74 years, 44.3% (N=66)
  • 75 to 84 years, 32.2% (N=48)
  • 85+ years, 7.4% (N=11).

Study 3:

• N=97
  • Less than 64 years, 3.1% (N=3)
  • 65 to 74 years, 17.5% (N=17)
  • 75 to 84 years, 53.6% (N=52)
  • 85+ years, 25.8% (N=25)

Other Relevant Demographics

	Study 1 N=193	Study 2 N=149	Study 3 N=97
Characteristic	%	%	%
Gender (female)	62.2	69.1	85.5
Living situation Lives alone Lives with spouse	33.7 56.5	37.8 48.6	NA NA
Marital status Married Widowed	60.1 26.9	NA NA	28.9 58.8
Education level Primary school or less High school graduate Post-secondary degree	5.2 26.4 37.8	8.7 30.9 34.2	17.5 22.7 33.0
Income less than $20,000/year	26.9	14	NA
General health rating Good to excellent Fair to poor	76.7 23.3	87.9 12.1	NA NA
Health-related QOL Health worse than last year Limited in vigorous activity Limited in moderate activity Limited walking stairs Limited bending/stooping Limited walking one block Limited ADL Receives help with ADL/IADL	20.8 76.7 44.1 46.1 42.7 26.9 9.3 43.5	NA 61.0 22.3 29.3 21.8 10.9 4.7 NA	NA NA NA NA NA NA NA 64.9
Dietitian rating Risk level         Low (rank one to four) Moderate (rank five to seven) High (rank eight to 10)	14.5 66.8 18.7	NA NA NA	NA NA NA

Location:

Guelph area of Ontario, Canada

 Proportion of Participants with Specific SCREEN II Item Scores that Indicate Increased Risk (Item Score <2), Study 1-3

Question	Study 1 (%)	Study 2 (%)	Study 3 (%)
More than 5-lb weight change	23	25.5	18.6
Unintentional weight change	4.7	4.7	14.4
Think excess weight	60.2	57.7	60.4
Think weight less than should be	8.3	4.7	4.2
Often/always skip meals	4.2	5.4	8.2
Limits foods	38.7	36.9	30.9
Fair/poor appetite	12.6	7.4	19.6
Less than or equal to three fruits and vegetables per day	33.0	43.0	8.2
Less than or equal to one meat and alternatives per day	22.5	39.6	12.4
Less than or equal to two meat and protein per day	43.9	55.0	28.9
Tree to four cups fluid per day	23.8	28.4	4.1
Sometimes/often swallow difficulty	17.5	15.4	13.4
Sometimes/often chewing difficulty	19.6	10.7	19.6
Sometimes/often use meal replacements	14.7	24.8	3.0
Sometimes/never eat with others	33.0	42.3	55.7
Prepares own meal	59.2	70.5	67.0
Shares meal prep with others	14.7	12.8	9.3
Other cooks	26.2	16.8	23.7
Sometimes cooking is a chore	29.6	30.0	23.7
Usually cooking a is a chore	9.5	7.0	15.5
Dissatisfied with others' cooking	0.5	0	1.0
Often/always difficulty with groceries	12.6	15.4	4.1
TOTAL SCREEN II SCORE
	Median	95% Confidence Interval
Study 1 sample	51.0	48.28, 50.37
Study 2 sample	49.0	47.27, 49.37
Study 3 sample	52.0	50.07, 52.41

 

Other Findings

Study 1:

• Seniors recruited from the geriatrician's clinic had significantly higher dietitian risk rating (7.3±6.8 vs. 5.7±1.5, T=-8.0, P=0.000) and lower SCREEN II scores (45.9±7.0 vs. 50.8±6.9, T=4.46, P=0.000) than seniors recruited from the community
• Participants who perceived their health to be excellent had significantly higher SCREEN II scores than those who considered their health to be poor (53.7±4.3 vs. 42.4±8.0, F=12.2, P=0.000)
•  SCREEN II was significantly correlated with the dietitian nutritional risk rating (P=-0.62, P=0.01)
• The ROC curve using the median dietitian risk rating 96.0 on a scale of one to 10 indicated that SCREEN II was an adequate index for comparison to this criterion
• The AUC was 82% and a score less than 54 was considered a good cut-point with a sensitivity of 84%, specificity of 62%, positive predictive value of 85% and a negative predictive value of 61%
• The eight-item abbreviated SCREEN II was also an adequate index when compared with the criteria. The AUC was 79% and a score less than 43 out of a possible 48 total had a sensitivity of 84%, specificity of 58%, positive predictive value of 83% and a negative predictive value of 59%
• A 'high-risk' category (ranking eight to 10 by dietitian risk rating) was also determined. The AUC was 82% and the best cutoff was less than 50 (SE=86%, SP=66%)
• A similar lower cut-point (score less than 38) for the abbreviated version was also determined (SE=77%, SP=64%, AUC=78%).

Study 2:

• Total SCREEN II score for test and retest were significantly and highly correlated [ICC=0.83, 95% CI (0.75, 0.88)]
• The test-retest reliability of the abbreviated SCREEN II is similar [ICC=0.84, 95% CI (0.79, 0.89)].

Study 3:

• Intra-class correlations indicate that inter-rater [ICC=0.83, 95% CI (0.74, 0.89)] and intra-rater reliability [ICC 0.90, 95% CI (0.80, 0.95)] were adequate for the full index
• Abbreviated SCREEN II had slightly lower reliability [ICC=0.79, inter-rater, ICC=0.85 intra-rater)
• Individual SCREEN II items were compared for test-retest, and inter-rater and intra-rater reliability
  • There was a range of ICC based on these analyses
  • There was no consistency among the three types of reliability testing for items with low correlations (ICC<0.5) suggesting that no specific items warranted removal or rewriting.

• Based on dietitian rankings for the validation sample, nutritional risk based on a detailed nutritional assessment and clinical judgement is prevalent in community-living seniors
• SCREEN II has a higher sensitivity, specificity and reliability than SCREEN I
• Recommended cut-off of less than 54 on the SCREEN II be used for health promotion and nutrition education efforts
• Recommended cut-off of less than 50 on the SCREEN II be used for a screening process
• SCREEN II is valid and reliable.

Government:

Canadian Institutes of Health Research

• A relatively well-conducted set of studies to construct validate and test the reliability of SCREEN II
• More information on exclusion criteria would be desired
• More information on the timing of events in Study 3 would be helpful
• Information on the ethnicity of the participants would enhance the validity, reliability and applicability of the data.