- Click here for explanation of classification scheme.

To design, pilot and evaluate a Nutrition Screening Tool (NST) based on the British Association of Parenteral and Enteral Nutrition (BAPEN4) (four questions and two measurements proposed as the minimum required for identifying patients with nutritional problems and possibly requiring referral for specialist nutrition advice).

• Study 1 (development): Patients admitted consecutively to the acute medical and elderly care wards at St. Thomas' Hospital in spring, 1998
• Study 2 (validity): Patients admitted consecutively to the acute medical and elderly care wards at St. Thomas' Hospital in spring, 1999 
• Study 3 (reliability): Patients on one acute medical admissions ward. 

All patients were included during a select time period, although some did not complete the study.

Recruitment

Not described in detail. 

 

Design

• Study 1 (development): Record the number of patients with positive responses to the BAPEN4, conduct a full nutritional assessment by an experienced dietitian to classify patients' nutritional risk, and use these results to develop a scoring system for a screening tool based on the BAPEN4
• Study 2 (validity): Nurse assessed patients using the developed nutrition screening tool (NST) and results were compared with a full nutrition assessment conducted by a dietitian
• Study 3 (reliability): Three nurses assessed the same patients using the developed NST.

Blinding Used 

• In Study 2 (validity), the dietitian and the nurses were blinded to the results of each other's assessments 
• In Study 3 (reliability), each nurse was blinded to the results obtained by the other nurses.

 

Intervention 

• Study 1 (development):
  • An experienced dietitian completed the BAPEN4 and full nutritional assessment within 72 hours of admission
    • Patients were weighed on portable scales (Soehnle Ltd., Germany) unless clinical condition dictated the use of ward chair scales. Both usual weight (obtained from patient or other source) and current weights were used to calculate percentage weight change.
    • Height was measured using a portable stadiometer (Leicester height measure), or estimated from demi-span measurement if height could not be measured
    • Measured height and weight were used to calculate body mass index (BMI) 
    • Mid-arm circumference (MAC) and triceps skinfold thickness (TSF) were measured according to published techniques and were used to calculate mid-arm muscle circumference (MAMC). These measurements were compared with gender- and age-specific standards.
    • Dietary intake (current and pre-illness) was assessed using diet history
    • Severity and duration of any gastro-intestinal symptoms was discussed
    • Examined for evidence of edema
    • Medical notes were reviewed for clinical condition
    • Nursing wound assessment form was reviewed to identify patients admitted with pressure sores
  • After the assessment, each patient was categorized by the dietitian as high, medium or low nutritional risk
  • Results from the assessment were summarized and an iterative process was used to develop a scoring system that would maximize the number of patients assigned to the correct categories, according to the dietitian’s assessment of risk
  • The initial draft of the screening tool was piloted by nurses on 45 patients on the medical, elderly care and respiratory wards at St. Thomas’ Hospital. The draft was reviewed by staff involved and amended based on comments. (Color-coded BMI charts were provided to reduce calculation errors, weight loss question was simplified to reported recent unintentional weight loss with no attempt to quantify the amount lost, by-mouth intake question was simplified to record poor intake without quantifying, MAC vs. current weight as an indicator of nutritional status, metric conversion charts were provided to avoid conversion errors, grade three to four pressure ulcers were incorporated into scoring system). 
  • The final tool was designed for use within 72 hours of admission and weekly throughout a patient’s hospital stay. It included space for up to five assessments on one sheet and nurses found it took only 34 minutes and was easy to use.  
  • Low-risk patients were monitored weekly throughout hospital stay; medium-risk patients were monitored daily and high-risk patients were discussed by the multi-disciplinary team and referred to the ward dietitian
  • A training package was designed for use prior to the introduction of the screening tool (included brief background to hospital malnutrition, how to use the screening tool and how to measure MAC). Written guidelines for completion of the NST and measurement of MAC were also prepared.

• Study 2 (validity)

  • Dietitian did full nutritional assessment on all patients within 72 hours of admission. For each patient, a nurse completed the NST also within 72 hours of admission.
  • Study 3 (reliability): Three nurses independently assessed 33 consecutive patients using the NST. 

 

Statistical Analysis

 

• Study 2 (validity)
  • Patients were assigned to low, medium or high risk for both the NST and the dietitian’s assessment
  • For the NST, patients were classified according to the score (zero to one, low; two to three, medium; four to five, high)
  • For the dietitian’s assessment, patients were classified according to the criteria used in the development phase 
  • An Excel database was constructed and statistical analysis was completed using SPSS (version 9.0)
  • Kappa statistics (a chance-corrected measure of agreement) were used to compare the agreement between the NST and the dietitian’s assessment
  • The dietitian’s assessment was considered to be the standard for all statistical analyses
  • All patients were assessed by the same dietitian and categorization was standardized the same as it was in Study 1.

• Study 3 (reliability)
  • Sample size was calculated according to the method described by Walter et al. 
  • With three assessors it was required that the NST be completed on 26 subjects to have an 80% power at P<0.05, of showing a difference in kappa between 0.6 and 0.8.

 

 

Timing of Measurements

• Study 1 (development) measurements were taken within 72 hours of admission
• Study 2 (validity) measurements were taken within 72 hours of admission
• Study 3 (reliability) unclear.

Dependent Variables

• Variable 1: Nutrition risk
  • High risk: BMI less than 18.5kg/m², or had experienced more than or equal to 10% weight loss in the previous three to six months or reported decreased dietary intake in the previous three to six months with accompanying weight loss exceeding 5%, nothing by mouth for more than five days for clinical reasons
  • Medium risk: If they had a BMI from 18.5 to 20.0kg/m² and had some recent weight loss or had experienced 5% to 10% recent weight loss or poor recent intake with minimal (less than 5%) weight loss. If, however, weight loss was likely to be ongoing, the patient was categorized as high risk. Patients with TSF, MAC, or MAMC less than fifth percentile were considered to be at nutritional risk (medium risk if there was no recent weight change and high risk if there had been significant recent weight change). 
  • All other patients were classified as low risk
• Variable 2: Inter-rater reliability (measured by comparing agreement between three nurses using the NST).

Independent Variables

• BAPEN4 results (recent unintentional weight loss, recent poor intake, usual weight, height)
• Full nutrition assessment (weight, height, BMI, usual weight, current weight, MAC, TSF, MAMC, dietary intake both current and pre-illness, nothing by mouth, severity and duration of gastro-intestinal symptoms, presence of edema, review of clinical condition and nursing wound assessment form).

 

Initial N

• Study 1: 127
• Study 2: 166
• Study 3: 33.

Attrition

Due to D/C, death, absent from ward during measurement time, unable or unwilling to comply, or NST not completed.

• Study 1: 27
• Study 2: 66
• Study 3: Seven.

Final N

• Study 1: 100
• Study 2: 100
• Study 3: 26.

Demographics

	Study 1	Study 2	Study 3
Males	59	63	15
Females	41	37	11
Age (years)	59.6 (19.4)	59.4 (16.6)	61.4 (16.9)

	Study 1	Study 2	Study 3
Body mass index (kg/m2)	25.2 (5.4)	26.2 (6.3)	24.0 (4.3)
Values for age and BMI are mean (SD); values for length of stay are medium (interquartile range).

Location

St. Thomas' Hospital NHS Trust, London, UK.

 

 

Study 1 (Development)

• Patients identified by the dietitian as at nutritional risk are highly likely to report recent unintentional weight loss and poor recent dietary intake
• Three high-risk patients did report no recent weight loss, including two who also reported no change in dietary intake; however, all three had BMI less than 18.5 kg/m² and MAC less than fifth percentile, indicating long-term, chronic undernutrition
• Edema was present in 11 (27%) of the medium-risk and high-risk patients and complicating the interpretation of weight and BMI (therefore it was addressed in the training package).

Study 2 (Validity)

• There was complete agreement between the dietitian’s assessment and the NST category in all the high-risk patients, 86% agreement of the low-risk patients and 59% of the medium risk patients
• Only two (3%) of the low-risk patients were classified as high risk by the NST, and none of the high-risk patients were classified as low risk by the NST
• The kappa value obtained using the contingency table (κ=0.72, 95% CI 0.59 to 0.84) indicates good agreement between the NST and the dietitian’s assessment.

Study 3 (Reliability)

• All three nurses were in agreement in 19 (73%) patients
• Kappa statistic showed reasonable inter-rater reliability (mean κ=0.66) with wide 95% confidence intervals.

The present NST was designed for use by nurses, doctors, health care assistants and other health care professionals, with little or no specialist knowledge of nutrition. It can potentially be used across a variety of specialties and in different clinical settings including hospital in-patients, outpatient clinics and GP clinics. Further studies are planned to evaluate its use in these settings.

Future work on the present NST includes refining the action plan to ensure patients are appropriately referred to other relevant health care professionals, e.g., speech and language therapists, occupational therapists and physiotherapists.

Finally, it is likely an NST can only be effective if it is completed on every patient within 72 hours of admission and regularly throughout hospital stay. To address this issue, regular audits of NST completion and appropriate interpretation of the action plan are being undertaken.

Government:

Research Training Fellowship Grant from the London Regional NHS Executive funded lead author

Only grade three to four pressure ulcers were considered for high risk.

Only high-risk patients were referred to an RD.

Patients were rescreened weekly by nursing staff; typically nursing staff in the US do not rescreen for nutritional risk but do track certain aspects of nutritional risk on a regular basis.

Many of the nuances of nutrition screening were covered during the training session that was only briefly described in the article. This would have directly affect inter-rater reliability and the effectiveness of the tool.