A simple dietary assessment tool to monitor food intake of hospitalized adult patients

Figures & data

Figure 1 Pictorial Dietary Assessment Tool (PDAT).

Notes: Mealtime: breakfast/lunch/dinner. For each level of intake, there are two values given for energy and protein (eg, 390/513 kcal and 23/25 g if 100% has been consumed). These two values represent the energy and protein contents for soft texture food (lower value) and normal texture food (higher value).

Table 1 Characteristics of health care staff and patients

Characteristics	Health care staff (n=37)	Patients (n=67)
Age (years), mean ± SD	40±6.6	44±15.4
Sex, n (%)
Female	33 (89.2)	43 (64.2)
Male	4 (10.8)	24 (35.8)
Background of health care staff, n (%)
Nurses	7 (18.9)
Dietitians	8 (21.6)
Serving assistants	22 (59.5)
Education level, n (%)
Middle (high school)	21 (56.8)
High (diploma, bachelor)	16 (43.2)
Years of working, mean ± SD	19±9.8
Type of diet, n (%)
Normal		17 (25.4)
Modified texture diet		14 (20.9)
Therapeutic diet		36 (53.7)
Wards, n (%)
Internal medicine		34 (50.7)
Neurology		21 (31.3)
Surgery		12 (17.9)

Figure 2 Bland-Altman of the difference in energy intake (kcal) between PDAT and weighed food values.

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Figure 3 Bland-Altman of the difference in protein intake (g) between PDAT and weighed food values.

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Figure 4 Bland-Altman of the difference in carbohydrate intake (g) between PDAT and weighed food values.

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Figure 5 Bland-Altman of the difference in fat intake (g) between PDAT and weighed food values.

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Table 2 Macronutrient intake estimated by the Pictorial Dietary Assessment Tool (PDAT) compared to food weighing (weighed)

Macronutrients	Breakfast (n=67)				Lunch (n=63)				Average (n=130)
	PDAT (mean ± SD)	Weighed (mean ± SD)	ra	P- valueb	PDAT (mean ± SD)	Weighed (mean ± SD)	ra	P-valueb	PDAT (mean ± SD)	Weighed (mean ± SD)	ra	P-valueb
Energy (kcal)	299±156	286±148	0.987^*	0.064	291±171	297±168	0.878^*	0.770	295±163	292±158	0.937^*	0.497
Protein (g)	14.0±7.7	14.1±8.0	0.941^*	0.385	13.8±8.0	14.0±7.9	0.916^*	0.515	13.9±7.8	14.1±8.0	0.896^*	0.521
Carbohydrate (g)	47.2±23.5	46.9±24.5	0.892^*	0.818	46.2±18.3	46.5±19.9	0.935^*	0.800	46.1±21.4	46.7±22.3	0.912^*	0.431
Fat (g)	7.5±3.25	7.5±3.25	0.965^*	0.931	7.2±2.8	7.3±2.9	0.904^*	0.282	7.4±3.06	7.4±3.10	0.952^*	0.377

Notes:

a Pearson’s correlation test (*P<0.01);

b paired t-test.

Abbreviation: SD, standard deviation.

Table 3 Accuracy of macronutrient intakes estimated by the PDAT in comparison with food weighing

Macronutrients	Accuracy	Bias percentage	P-valuec
Energy
• P15a	93.9	6.2	1
• P30b	93.9	6.2	0.727
Protein
• P15a	90.0	10.0	1
• P30b	99.2	0.8	1
Carbohydrate
• P15a	91.5	8.5	0.549
• P30b	90.0	10.0	0.092
Fat
• P15a	94.6	5.4	0.453
• P30b	90.8	9.2	0.039^*

Notes:

a Percentage of estimates within 15% of food weighing (gold standard);

b percentage of estimates within 30% of food weighing (gold standard);

c McNemar test (^*P<0.05).

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Table 4 Health care staff estimates of percentage food consumed by patients compared to food weighing method

Level of food intake using food weighing method	Number of patients eating each levela	Distribution of estimates using PDAT						Estimates correctly reported (%)	Agreement beyond chance (k)
		0%	25%	50%	75%	80%	100%
Staple food (n=130)
0%	12	12b	0	0	0	0	0	100	0.809*
25%	18	1	15b	2	0	0	0	83
50%	25	0	2	20b	3	0	0	80
75%	15	0	0	1	9b	5	0	60
80%	19	0	0	0	2	16b	1	84
100%	41	0	0	1	0	2	38b	93
Animal source protein (n=129)
0%	13	12b	1	0	0	0	0	92	0.805*
25%	7	0	4b	3	0	0	0	57
50%	16	0	2	12b	1	1	0	75
75%	8	0	0	1	4b	3	0	50
80%	10	0	0	0	1	9b	0	90
100%	75	0	0	0	0	3	72b	96
Non-animal source protein (n=120)
0%	35	33b	2	0	0	0	0	94	0.735*
25%	19	6	10b	3	0	0	0	53
50%	15	0	2	10b	1	0	0	67
75%	5	0	0	0	3b	2	0	60
80%	5	0	0	1	0	2b	2	40
100%	41	0	0	0	1	2	38b	93

Notes:

a Based on weighed plate waste;

b number of accurate estimates at this consumption level;

* significant (P<0.001).

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Table 5 Values of sensitivity and specificity of PDAT according to different levels of plate waste (food weighing)

Plate waste	True		False		Sensitivity	Specificity	Area under the curve
	positives	negatives	positives	negatives
Staple food (n=130)
≤50%	97	3	27	3	97	90	0.935
≤75%	118	0	12	0	100	100	1
Animal source protein (n=129)
≤50%	107	5	15	2	98.2	75	0.866
≤75%	115	3	10	1	99.1	76.9	0.880
Non-animal source protein (n=120)
≤50%	62	7	47	4	93.9	87	0.904
≤75%	83	4	2	31	97.6	88.6	0.931

Abbreviation: PDAT, Pictorial Dietary Assessment Tool.

Table 6 Inter-rater reliability analysis of estimated nutrient intake using PDAT among health care staff and comparison with food weighing

Macronutrients	Estimated nutrient intake by different health care staff (mean ± SD)			Intraclass correlation coefficient (95% CI)
	Nurses (n=7)	Dietitians (n=8)	Serving assistants (n=22)
Energy (kcal)	299.8±163.6	295.1±162.8	299.5±163.8	0.96 (0.94–0.97)
Protein (g)	14.1±7.8	13.8±7.9	13.8±8.0	0.91 (0.88–0.91)
Carbohydrate (g)	47.3±21.4	46.1±21.5	47.0±21.7	0.95 (0.94–0.96)
Fat (g)	7.2±3.0	7.4±3.1	7.2±3.1	0.91 (0.88–0.91)

Abbreviations: PDAT, Pictorial Dietary Assessment Tool; SD, standard deviation; CI, confidence interval.