European Journal of Clinical Nutrition (2015) 69, 187–192 © 2015 Macmillan Publishers Limited All rights reserved 0954-3007/15 www.nature.com/ejcn
ORIGINAL ARTICLE
Handgrip strength measurement as a predictor of hospitalization costs RS Guerra1,2,3, TF Amaral2,4, AS Sousa3,4, F Pichel3, MT Restivo2, S Ferreira5 and I Fonseca3 BACKGROUND: Undernutrition status at hospital admission is related to increased hospital costs. Handgrip strength (HGS) is an indicator of undernutrition, but the ability of HGS to predict hospitalization costs has yet to be studied. OBJECTIVE: To explore whether HGS measurement at hospital admission can predict patient's hospitalization costs. SUBJECTS/METHODS: A prospective study was conducted in a university hospital. Inpatient's (n = 637) HGS and undernutrition status by Patient-Generated Subjective Global Assessment were ascertained. Multivariable linear regression analysis, computing HGS quartiles by sex (reference: fourth quartile, highest), was conducted in order to identify the independent predictors of hospitalization costs. Costs were evaluated through percentage deviation from the mean cost, after adjustment for patients' characteristics, disease severity and undernutrition status. RESULTS: Being in the first or second HGS quartiles at hospital admission increased patient's hospitalization costs, respectively, by 17.5% (95% confidence interval: 2.7–32.3) and 21.4% (7.5–35.3), which translated into an increase from €375 (58–692) to €458 (161–756). After the additional adjustment for undernutrition status, being in the first or second HGS quartiles had, respectively, an economic impact of 16.6% (1.9–31.2) and 20.0% (6.2–33.8), corresponding to an increase in hospitalization expenditure from €356 (41–668) to €428 (133–724). CONCLUSIONS: Low HGS at hospital admission is associated with increased hospitalization costs of between 16.6 and 20.0% after controlling for possible confounders, including undernutrition status. HGS is an inexpensive, noninvasive and easy-to-use method that has clinical potential to predict hospitalization costs. European Journal of Clinical Nutrition (2015) 69, 187–192; doi:10.1038/ejcn.2014.242; published online 5 November 2014
INTRODUCTION Although this problem has been documented for a long time, a large number of patients are still undernourished when admitted to hospital.1,2 Moreover, nutritional status of a high proportion of patients worsens during hospital stay.3 There is also a considerable amount of scientific evidence showing that undernutrition, often under-recognized and untreated,1,4 is associated with higher hospitalization costs.5–8 Data from multivariable analysis from studies conducted in Portugal (2007) and in Singapore (2012) reveal that the cost of a patient at risk of undernutrition or undernourishment, respectively, is approximately 19.3–24% higher than for a patient not presenting any of these conditions with the same diagnosis-related group (DRG) code.5,6 After controlling for underlying conditions and treatments, data from 2003 to 2004 show that undernutrition was responsible for an increase in AU $1745 per hospital admission.7 It has been recently documented in Croatia that adult undernutrition costs over €97 million annually, accounting for 3.4% of the total Croatian national health-care budget.8 The higher hospitalization costs associated with undernutrition can be attributed to increased risk for infections 9,10 and pressure ulcers,10,11 higher complications rate12 and also longer hospital stay13,14 and the readmission rate.6 Moreover, undernutrition is known to diminish effectiveness of medical treatment15 and to contribute to delayed functional recovery during hospitalization.16
This information on the economic impact of undernutrition emphasizes the importance of identifying and treating undernourishment, as it has been shown that screening and nutritional treatment of undernourished patients can lower hospitalization costs.17–19 There is an increasing evidence that handgrip strength (HGS) is a reliable and attractive method used in clinical daily practice to identify undernutrition.20–22 Undernutrition is associated with impaired hand muscle function and HGS reflects early nutrition deprivation and repletion, before changes in body composition parameters can be detected.20 Moreover, HGS is the most frequently used indicator of muscle function for clinical purposes;20 it has been shown to be a reliable indicator of functional ability20 and also an outcome predictor.20,23–25 Dynamometers used to measure HGS are inexpensive and portable. In addition, the measurement of HGS is noninvasive, easy and quick to perform with no need of specialized professionals, reliable 26 and shows low intra- and inter-observers variability.26 Because of the budgetary constraints that health-care systems face today,2 hospitalization undernutrition-related costs are a key issue in health-care management. Therefore, it is of the utmost importance to seek a simple and reliable tool to predict hospitalization expenditure so that a timely and effective health-care plan can be implemented. This study was designed to explore the ability of HGS measurement to predict patient's hospitalization costs,
1 Departamento de Bioquímica, Faculdade de Medicina da Universidade do Porto, Porto, Portugal; 2UISPA-IDMEC, Faculdade de Engenharia da Universidade do Porto, Porto, Portugal; 3Unidade de Nutrição, Centro Hospitalar do Porto, Porto, Portugal; 4Faculdade de Ciências da Nutrição e Alimentação da Universidade do Porto, Porto, Portugal and 5 Serviço de Informação de Gestão, Centro Hospitalar do Porto, Porto, Portugal. Correspondence: RS Guerra, Departamento de Bioquímica, Faculdade de Medicina da Universidade do Porto, Al. Professor Hernâni, Porto 4200-319, Portugal. E-mail: [email protected] Received 02 June 2014; revised 16 September 2014; accepted 27 September 2014; published online 5 November 2014
Handgrip strength and hospitalization costs RS Guerra et al
188 considering the possible confounder effect of patients' characteristics, disease severity and undernutrition status. SUBJECTS AND METHODS Study population and design A prospective study was conducted in a Portuguese university hospital between July 2011 and June 2013. A consecutive sampling approach was used. From the daily list of patients admitted to each ward, those who met the inclusion criteria were invited to participate in the study. Patients were eligible to participate in the study if they were ⩾ 18 years old, Caucasian, with an expected length of stay 424 h, conscious, cooperative and able to provide written informed consent. Patients unable to perform the HGS technique, defined as unable to understand verbal instructions or having a condition in which the patient could not perform the technique correctly and also patients with critical illness, defined as failure of at least one vital organ27 and admitted to intensive care units, were excluded from the study. Pregnant women, individuals in isolation, those who were admitted for procedures that could put them in a critical situation and those with hemodynamic instability at the time of evaluation were also excluded. Therefore, inpatients from angiology and vascular surgery, cardiology, digestive, non-digestive and hepato-biliary surgeries, endocrinology, gastroenterology, internal medicine, nephrology, orthopedics, otolaryngology and urology wards were considered eligible for this study. Sample size for linear regression was calculated assuming an effect size of 0.15, a statistical power of 80% and a level of significance equal to 0.05. The minimum required sample size was 113 individuals. In order to have a representative sample of the hospitalized patients who met the inclusion criteria, data collection took place until the number of patients corresponded to twice the total number of beds of all the selected wards (n = 711). Therefore, this study sample had sufficient statistical power to identify the independent variables associated with hospitalization costs. Inpatients transferred from intensive care units to the hospital ward where data were collected and inpatients transferred from surgical to medical wards (n = 48) were not included in data analysis, given that these participants' hospitalization percentage of cost deviation was above the superior limit of 95% confidence interval (CI) of cost estimates. Given the likelihood of providing inaccurate information, patients with possible cognitive impairment (n = 13) and also patients with missing data (n = 13) were excluded from data analysis. The final study sample was composed of 637 participants.
Ethics This research was conducted according to the guidelines laid down in the Declaration of Helsinki and approved by the Institutional Review Board and the Ethics Committee of Centro Hospitalar do Porto. Written informed consent was obtained from all study participants.
Data collection Demographical data and clinical history were obtained by consulting the patient's clinical file. DRG codes were obtained from hospital records. The remaining data were collected within 72 h of admission to the hospital by two previously trained nutritionists using a structured questionnaire. The education level was evaluated by the number of completed schooling years and the following classes were created: 0–4, 5–12 and 412 school years. Marital status was categorized as single, married or in a common-law marriage, divorced and widowed. Disease severity was evaluated using the Charlson comorbidity index.28 Medical discharge diagnoses in the patient's clinical record were obtained by a previously trained interviewer and used to classify comorbidity conditions. The Charlson index takes into account the number and the seriousness of comorbid diseases, each scored from 0 to 6.28 Possibility of cognitive impairment was assessed using the Abbreviated Mental Test.29 This test consists of 10 questions, each scored one point if correct, with a cutoff score of 7 or 8 out of 10 discriminating between cognitive impairment and normality in older adults.29 Although the Abbreviated Mental Test was developed to discriminate cognitive impairment of older adults, it has also been applied in adults younger than 65 years.30 In the present study, a score o6 was considered denotative of cognitive impairment because it has shown the best combination of sensitivity and specificity in a mixed sample of adults and older adults.31 Participants' undernutrition status was evaluated by Patient-Generated Subjective Global Assessment.32 European Journal of Clinical Nutrition (2015) 187 – 192
Nondominant HGS was measured in kgf with a calibrated Jamar Hand Dynamometer (Sammons Preston, Bolingbrook, IL, USA), recommended by the American Society of Hand Therapists.33 Measurements were performed in a chair or on a bed, with the shoulder adducted and neutrally rotated, elbow flexed to an angle of 90º and the forearm and wrist in neutral position.34 Each participant performed three measurements with a 1 min pause between measurements and the maximum value was chosen as the HGS value.35 When the individual was unable to perform the measurement with the nondominant hand, the dominant hand was used (n = 53). Standing height (cm) was obtained with a metal tape (Rosscraft Innovations Incorporated, Surrey, Canada), with 0.1-cm resolution and a headboard.36 For the participants with visible kyphosis or when it was impossible to measure standing height (n = 312), recumbent height was obtained.37,38 Body weight (kg)36 was measured with a calibrated portable beam scale with 0.5-kg resolution. For participants on dialytic therapies (n = 19), dry body weight was used. When it was not possible to weigh a patient (n = 148), body weight was estimated from height and mid-upper arm circumference.39 Mid-upper arm circumference was measured with a metal tape (Rosscraft Innovations Incorporated, Surrey, Canada), with 0.1-cm resolution.40 The intra- and inter-observer technical error of measurement was obtained for all anthropometric measurements, respectively, in 17 and 18 individuals. Intra-observer error varied between 0.2 and 0.6%, and interobserver error varied between 0 and 1.4%, respectively, values considered acceptable for trained anthropometrists.41
Statistical analysis The Kolmogorov–Smirnov test was used to evaluate the normality of variable distribution. Continuous normally distributed variables were described as mean and s.d. and non-normal as median and interquartile range. Categorical variables were reported as frequencies. HGS was summarized using quartiles by sex, according to the cutoffs of sample distribution stratified by sex. Cutoffs were 12.0, 16.3 and 21.9 kgf for women and 25.8, 32.0 and 37.8 kgf for men. HGS varied between 1.0 and 35.1 kgf for women; median (interquartile range) was equal to 16.4 (9.6) kgf. For men, HGS values varied between 1.0 and 61.0 kgf, with a median (interquartile range) of 32.0 (12.0) kgf. Hospitalization cost was calculated for each inpatient on the basis of discharge DRG codes. The DRG system is used to calculate hospital reimbursements, with the amounts determined on the basis of a relative weight value. This weight value reflects the main diagnostic, surgical interventions, associated pathologies and complications, clinical procedures, medium length of hospital stay, patient age and sex and discharge destination. The information about DRG codes and its amounts was obtained from Portuguese Ministerial Directive number 839-A, 31 July 2009,42 used for data obtained between 2011 and 2012 and number 163, 24 April 2013, 43 which replaced the latter directive, was used for data obtained in 2013. The percentage of cost deviation was calculated from the difference between the cost of each inpatient and the cost of a patient with relative weight of one (€2396 for data obtained between 2011 and 2012; €2142 for data obtained in 2013). Percentage of cost deviation was summarized using quartiles according to the cutoffs of the sample distribution (−46.0, 1.3 and 49.0%). In order to identify variables associated with HGS and with the percentage of cost deviation, patients' demographic and clinical characteristics were compared across HGS quartiles and across the percentage of cost deviation quartiles using one-way Anova (or the Kruskal–Wallis test) for continuous variables and the Pearson χ2-test for categorical variables. Multivariable linear regression analysis using the stepwise method was performed to identify the independent variables of percentage of cost deviation. The following variables were included in the model: HGS quartiles, age, education, professional activity, marital status, hospitalization in medical or surgical wards, Abbreviated Mental Test and Charlson comorbidity index scores. As HGS is an indicator of the nutrition status,1,20–22 a multivariable analysis was conducted including all the mentioned variables and adjusting for undernutrition status. The interaction between hospitalization in medical or surgical wards and HGS quartiles was also tested. Results were considered significant when Po0.05. Statistical analyses were conducted using the Software Package for Social Sciences (SPSS) for Windows, version 21.0; SPSS, Inc, an IBM Company, Chicago, IL, USA. © 2015 Macmillan Publishers Limited
Handgrip strength and hospitalization costs RS Guerra et al
RESULTS The characteristics of the 637 patients (aged between 18 and 91 years old) stratified by HGS quartiles are presented in Table 1. Participants with lower HGS values were older, presented lower Abbreviated Mental Test and higher Charlson comorbidity index scores. A lower percentage of participants in work and higher proportions of widows, patients with ⩽ 4 years of schooling and admitted in medical wards, were observed in the lower quartiles of HGS compared with the higher quartiles of HGS. Moreover, a higher proportion of patients who were severely undernourished was observed in the three lower quartiles of HGS. In the two lower HGS quartiles, a positive percentage of cost deviation was found. Thus, participants with low HGS have a mean cost higher compared with the cost of a patient with relative weight of one. In contrast, the percentage of cost deviation was negative in the two higher quartiles of HGS, which means that participants with high HGS have a mean cost lower compared with the cost of a patient with relative weight of one. Patient's hospitalization costs within this sample varied between €393 and €20447, with a median (interquartile range) equal to €2428 (€3030). Participant's characteristics according to the percentage of cost deviation quartiles are presented in Table 2. Participants in the two lower quartiles of percentage of cost deviation were younger, single or divorced, had a professional activity and a higher number of schooling years. Otherwise, Table 1.
189 a higher proportion of inpatients admitted in medical wards was observed in the lower quartiles of percentage of cost deviation. Patients in the two lower quartiles of percentage of cost deviation presented better nutrition status by Patient-Generated Subjective Global Assessment, compared with the two higher quartiles. A higher proportion of patients with high HGS in the lower quartiles of percentage of cost deviation was also observed. The multivariable linear regression models computed to predict the percentage of cost deviation are presented in Table 3. In model 1, the predictors of higher hospitalization costs were advanced age, hospitalization in surgical wards and being in the first (the lowest) or in the second HGS quartiles. In model 2, it is shown that advanced age, hospitalization in surgical wards, being in the first (the lowest) or in the second HGS quartiles and severe undernutrition predicted higher percentage of cost deviation. The interaction between hospitalization in medical or surgical wards and HGS quartiles was tested, but no significant effect was found in any of the fitted models (P ⩾ 0.349). According to our estimation, and after controlling for possible confounders, being in the first or second HGS quartiles at hospital admission is related to increased odds of having higher hospitalization costs of between €375 (95% CI: €58–692) and €458 (95% CI: €161–756). After the additional adjustment for undernutrition status, being in the first or second HGS quartiles
Sample characteristics according to handgrip strength quartiles Handgrip strength quartilesa (n = 637)
Characteristic
1st (lowest) (n = 154)
2nd (n = 166)
3rd (n = 151)
4th (highest) (n = 166)
P
Sex, n (%) Women Men
72 (46.8) 82 (53.2)
72 (43.4) 94 (56.6)
73 (48.3) 78 (51.7)
73 (44.0) 93 (56.0)
Age (years), median (IQR)
65 (21)
63 (19)
57 (21)
47 (24)
o 0.001c
Having a professional activity, n (%)
25 (16.2)
35 (21.1)
49 (32.5)
82 (49.4)
o 0.001b
Education (years), n (%) 0–4 5–12 412
86 (56.2) 59 (38.6) 8 (5.2)
74 (44.8) 76 (46.1) 15 (9.1)
60 (40.0) 76 (50.7) 14 (9.3)
50 (30.1) 98 (59.0) 18 (10.8)
0.001b
Marital status, n (%) Single Married/common-law marriage Divorced Widow
23 99 12 20
Hospital wards, n (%) Medical Surgical
90 (33.1) 64 (17.5)
(14.9) (64.3) (7.8) (13.0)
22 107 13 24
(13.3) (64.5) (7.8) (14.5)
67 (24.6) 99 (27.1)
26 104 13 8
(17.2) (68.9) (8.6) (5.3)
64 (23.5) 87 (23.8)
0.790b
(19.9) (65.1) (10.8) (4.2)
0.032b
51 (18.8) 115 (31.5)
o 0.001b
33 108 18 7
Abbreviated Mental Test (score), median (IQR)
9 (2)
9 (1)
10 (1)
10 (1)
0.006c
Charlson comorbidity index (score), median (IQR)
2 (2)
2 (3)
2 (3)
1 (2)
o 0.001c
126 (75.9) 17 (10.2) 23 (13.9)
o 0.001b
PG-SGA, n (%) Not undernourished Moderate or suspected undernutrition Severe undernutrition
55 (37.7) 58 (35.7) 41 (26.6)
73 (44.0) 48 (28.9) 45 (27.1)
83 (55.0) 34 (22.5) 34 (22.5)
Body mass index, kg/m2, mean (s.d.)
25.7 (5.1)
26.6 (5.5)
26.0 (5.0)
27.2 (5.3)
Percentage of cost deviation, €, median (IQR)
23.6 (94.2)
31.9 (94.5)
− 10.8 (94.1)
− 12.2 (101.2)
0.052d o 0.001c
Abbreviations: IQR, interquartile range; PG-SGA, Patient-Generated Subjective Global Assessment. aCutoffs: 12.0, 16.3 and 21.9 kgf for women; 25.8, 32.0 and 37.8 kgf for men. bPearson χ2-test. cKruskal–Wallis test. dOne-way ANOVA.
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 187 – 192
Handgrip strength and hospitalization costs RS Guerra et al
190 Table 2.
Sample characteristics according to percentage of cost deviation quartiles Percentage of cost deviation quartiles (n = 637)
Characteristic
1st −509.1, −46.0 (n = 159)
2nd −45.9, 1.3 (n = 163)
3rd 1.4, 49.0 (n = 156)
4th 49.1, 88.3 (n = 159)
P
Sex, n (%) Women Men
76 (47.8) 83 (52.2)
73 (44.8) 90 (55.2)
64 (41.0) 92 (59.0)
77 (48.4) 82 (51.6)
Age (years), median (IQR)
53 (29)
52 (23)
63 (19)
62 (16)
o 0.001b
Having a professional activity, n (%)
61 (38.4)
64 (39.3)
33 (21.2)
33 (20.8)
o 0.001a
Education (years), n (%) 0–4 5–12 412
52 (32.9) 87 (55.1) 19 (12.0)
62 (38.0) 86 (52.8) 15 (9.2)
74 (47.7) 67 (43.2) 14 (9.0)
82 (51.9) 69 (43.7) 7 (4.4)
0.008a
Marital status, n (%) Single Married/common-law marriage Divorced Widow
33 102 16 8
(20.8) (64.2) (10.1) (5.0)
34 100 17 12
(20.9) (61.3) (10.4) (7.4)
23 103 12 18
(14.7) (66.0) (7.7) (11.5)
0.536a
(8.8) (71.1) (6.9) (13.2)
0.019a
41 (25.8) 118 (74.2)
o 0.001a
14 113 11 21
Hospital wards, n (%) Medical Surgical
86 (54.1) 73 (45.9)
77 (47.2) 86 (52.8)
68 (43.6) 88 (56.4)
Abbreviated mental test (score), median (IQR)
10 (1)
10 (1)
10 (1)
9 (1)
0.526b
2 (2)
2 (2)
2 (2)
2 (3)
0.080b
Charlson comorbidity index (score), median (IQR) PG-SGA, n (%) Not undernourished Moderate or suspected undernutrition Severe undernutrition
102 (64.2) 32 (20.1) 25 (15.7)
Body mass index, kg/m2, mean (s.d.)
26.3 (5.3)
26.0 (5.2)
27.0 (5.5)
26.4 (5.1)
0.429c
Handgrip strength, kgf, median (IQR)
25.0 (17.0)
24.0 (15.9)
24.0 (15.8)
22.0 (16.6)
0.072b
Handgrip strength quartilesd, kgf, n (%) 1st (lowest) 2nd 3rd 4th (highest)
34 35 38 52
(21.4) (22.0) (23.9) (32.7)
93 (57.1) 46 (28.2) 24 (14.7)
35 33 48 47
(21.5) (20.2) (29.4) (28.8)
73 (46.8) 40 (25.8) 43 (27.6)
41 45 32 38
(26.3) (28.8) (20.5) (24.4)
72 (45.3) 36 (22.6) 51 (32.1)
44 53 33 29
(27.7) (33.3) (20.8) (18.2)
o 0.001a
0.021a
Abbreviations: IQR, interquartile range; PG-SGA, Patient-Generated Subjective Global Assessment. aPearson χ2-test. bKruskal–Wallis test. cOne-way ANOVA. d Cutoffs: 12.0, 16.3 and 21.9 kgf for women; 25.8, 32.0 and 37.8 kgf for men.
had, respectively, an economic impact of between €356 (95% CI: €41–668) and €428 (95% CI: €133–724). DISCUSSION Findings from the present study clearly demonstrate that patients with lower HGS at hospital admission have higher hospitalization costs. This is, as far as we are concerned, the first study to show that HGS at hospital admission predicts deviations from the hospitalization cost of a standard patient with relative weight of one. HGS is considered as an indicator of functional and nutrition status.1,20–22 It is therefore important to highlight that low HGS was a significant predictor of higher hospitalization costs, even after controlling for undernutrition status. It was also shown in the model not adjusted for undernutrition status that having low HGS, advanced age and being hospitalized in surgical wards independently predicted higher hospitalization costs. Moreover, an updated European Journal of Clinical Nutrition (2015) 187 – 192
estimate of hospitalization costs associated with patient's severe undernutrition according to Patient-Generated Subjective Global Assessment is provided, highlighting the economic burden of undernutrition at hospital setting. Examining the explained variation assessed by the adjusted coefficients of multiple determination (adjusted R2), it is also noteworthy that the two models share a similar explanatory power in predicting hospital costs. Undernutrition, after the adjustment for several confounders, represented an independent increase in costs of 19.3% in Portugal in 2007,5 whereas in Singapore the average cost of hospitalization was 24% higher for undernourished patients in 2012.6 Compared with these previous studies, low HGS in the present sample represented a similar independent increase in hospitalization costs between 16.6 and 21.4%. Exclusion criteria should be acknowledged because the sample restriction to patients able to provide written informed consent limits the generalization of study results for hospitalized patients who are not in this situation. © 2015 Macmillan Publishers Limited
Handgrip strength and hospitalization costs RS Guerra et al
191 Table 3. Multivariable linear regression models for prediction of hospitalization cost deviationa,b Regression coefficient Model 1d Age (years) Hospital wards (surgical vs medical)
0.80 27.3
95% CI
0.43–1.17 o0.001 15.8–38.8 o0.001
Handgrip strength quartilese (reference: 4th quartile) 1st (lowest) 17.5 2.7–32.3 2nd 21.4 7.5–35.3 3rd 0.07 − Model 2f PG-SGA (reference: not undernourished) Moderate/suspected undernutrition Severe undernutrition Age (years) Hospital wards (surgical vs medical)
0.06 24.7 0.70 24.4
Pc
−
0.020 0.003 0.159
0.151
11.3–38.2 o0.001 0.33–1.07 o0.001 16.1–38.8 o0.001
Handgrip strength quartilese (reference: 4th quartile) 1st (lowest) 16.6 1.9–31.2 2nd 20.0 6.2–33.8 3rd 0.06 −
0.027 0.005 0.209
Abbreviations: CI, confidence interval; PG-SGA, Patient-Generated Subjective Global Assessment. aCalculated as a percentage from the difference between the cost of each inpatient and the cost of a patient with relative weight of one. bVariables included the following: Model 1, HGS quartiles (fourth HGS quartile, highest, was used as reference), age, education (having 412 schooling years was used as reference), professional activity (having a professional activity was used as reference), marital status (being married or in common-law marriage was used as reference), hospitalization in medical or surgical wards (hospitalization in medical wards was used as reference), the Abbreviated Mental Test score, the Charlson comorbidity index score. Model 2, all the aforementioned variables and undernutrition status by PG-SGA (‘not undernourished’ was used as reference). cP-value by linear regression analysis. dAdjusted R2 = 0.081. eCutoffs: 12.0, 16.3 and 21.9 kgf for women; 25.8, 32.0 and 37.8 kgf for men. fAdjusted R2 = 0.099.
Despite the higher costs that undernutrition implies, identification of undernourished patients and proper nutritional treatment can reduce hospitalization costs.17–19 A study from 2005 showed that the incremental cost of 1 less day in hospital for undernourished patients due to nutritional treatment was €76, whereas the mean cost of 1 day in hospital was €476 for university hospitals and €337 for peripheral hospitals.17 This study stated that the incremental costs of undernutrition screening and treatment to reduce the length of hospital stay by 1 day were higher in undernourished patients who had HGS higher than the standard compared with those of undernourished patients with HGS lower than the standard.17 This is probably because frail patients, that is, patients simultaneously undernourished and with low HGS would benefit more from nutritional intervention.17 In a study conducted among patients subjected to liver transplant, the assessment of preoperative nutrition status included HGS, but no single or combination of parameters predicted global resource utilization, which is an index of cost.44 Patients identified as undernourished were given nutritional support, which can negate the adverse effects of undernutrition on post-transplant outcomes.44 Other explanations for this finding are that only five individuals (9.4% of study sample) were severely undernourished and liver transplant is a high-cost situation where HGS could have a lower predictive value.44 Some strengths of the present study can be enumerated. The association between HGS and hospitalization cost was adjusted for potential confounders, including undernutrition status. Also, © 2015 Macmillan Publishers Limited
the study sample is composed of a large number of inpatients within a varying age range (18–91 years old) from a variety of hospital wards and therefore ensuring a wide spectrum of diagnoses and relevant pathologies. Moreover, present sample HGS values and quartile cutoffs are comparable to other samples of inpatients.45,46 These aspects strengthen the external validity of the present study results for other hospitalized patients. The DRG system was shown to underestimate the real hospitalization costs.47 Moreover, only direct hospitalization costs are considered, whereas indirect costs, such as productivity loss due to time off work or societal costs,48 are not. Nevertheless, this methodology has been used extensively to estimate undernutrition costs, 5,6,8 as it allows to cluster patients with a variety of diagnoses and procedures into diagnostic groups. Cases within a group will have a similar level of complexity, and their treatment is expected to have similar costs.49 For these reasons, the DRG cost estimation methodology was used in this study. Until now, the ability of HGS measurement at admission to predict patient's hospitalization cost has never been assessed, justifying the current study. Results first show that low HGS is associated with higher hospitalization costs, considering the possible confounder effect of other factors such as hospitalization in surgical wards and undernutrition. Patients with a low HGS at hospital admission have an independent increase in hospitalization costs ranging from 16.6% (95% CI: 1.9–31.2%) to 21.4% (95% CI: 7.5–35.3%), representing €356 (95% CI: €41–668) and €458 (95% CI: €161–756). The early identification of patients with low HGS offers the opportunity to treat them, namely by nutrition intervention, which can improve their clinical progress and outcome and possibly reduce hospitalization costs. Future studies should compare HGS ability in predicting costs with that of other undernutrition measures and tools. Also, experimental studies should evaluate the cost-effectiveness of nutritional treatment of patients who are identified with low HGS values at hospital admission. HGS is an inexpensive, noninvasive and easy-to-use method that has clinical potential to predict hospitalization costs. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank Centro Hospitalar do Porto and all ward directors for facilitating the data collection. Rita S Guerra as a PhD student received a scholarship from FCT—Fundação para a Ciência e a Tecnologia under the project (SFRH/BD/61656/2009).
REFERENCES 1 White JV, Guenter P, Jensen G, Malone A, Schofield M. Consensus statement: Academy of Nutrition and Dietetics and American Society for Parenteral and Enteral Nutrition: characteristics recommended for the identification and documentation of adult malnutrition (undernutrition). JPEN J Parenter Enteral Nutr 2012; 36: 275–283. 2 Stratton RJ, Elia M. Encouraging appropriate, evidence-based use of oral nutritional supplements. Proc Nutr Soc 2010; 69: 477–487. 3 Rasmussen HH, Holst M, Kondrup J. Measuring nutritional risk in hospitals. Clin Epidemiol 2010; 2: 209–216. 4 Tappenden KA, Quatrara B, Parkhurst ML, Malone AM, Fanjiang G, Ziegler TR. Critical role of nutrition in improving quality of care: an interdisciplinary call to action to address adult hospital malnutrition. J Acad Nutr Diet 2013; 113: 1219–1237. 5 Amaral TF, Matos LC, Tavares MM, Subtil A, Martins R, Nazare M et al. The economic impact of disease-related malnutrition at hospital admission. Clin Nutr 2007; 26: 778–784. 6 Lim SL, Ong KC, Chan YH, Loke WC, Ferguson M, Daniels L. Malnutrition and its impact on cost of hospitalization, length of stay, readmission and 3-year mortality. Clin Nutr 2012; 31: 345–350.
European Journal of Clinical Nutrition (2015) 187 – 192
Handgrip strength and hospitalization costs RS Guerra et al
192 7 Rowell DS, Jackson TJ. Additional costs of inpatient malnutrition, Victoria, Australia, 2003-2004. Eur J Health Econ 2011; 12: 353–361. 8 Benkovic V, Kolcic I, Ivicevic Uhernik A, Vranesic Bender D, Oreb I, Stevanovic R et al. The economic burden of disease-related undernutrition in selected chronic diseases. Clin Nutr 2014; 33: 689–693. 9 Schneider SM, Veyres P, Pivot X, Soummer AM, Jambou P, Filippi J et al. Malnutrition is an independent factor associated with nosocomial infections. Br J Nutr 2004; 92: 105–111. 10 Fry DE, Pine M, Jones BL, Meimban RJ. Patient characteristics and the occurrence of never events. Arch Surg 2010; 145: 148–151. 11 Banks MD, Graves N, Bauer JD, Ash S. The costs arising from pressure ulcers attributable to malnutrition. Clin Nutr 2010; 29: 180–186. 12 Correia MI, Waitzberg DL. The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clin Nutr 2003; 22: 235–239. 13 Leandro-Merhi VA, de Aquino JL, Sales Chagas JF. Nutrition status and risk factors associated with length of hospital stay for surgical patients. JPEN J Parenter Enteral Nutr 2011; 35: 241–248. 14 Pavic T, Ljubicic N, Stojsavljevic S, Krznaric Z. Nutritional screening model in tertiary medical unit in Croatia. Ann Nutr Metab 2012; 61: 65–69. 15 Jensen GL, Mirtallo J, Compher C, Dhaliwal R, Forbes A, Grijalba RF et al. Adult starvation and disease-related malnutrition: a proposal for etiologybased diagnosis in the clinical practice setting from the International Consensus Guideline Committee. JPEN J Parenter Enteral Nutr 2010; 34: 156–159. 16 Covinsky KE, Martin GE, Beyth RJ, Justice AC, Sehgal AR, Landefeld CS. The relationship between clinical assessments of nutritional status and adverse outcomes in older hospitalized medical patients. J Am Geriatr Soc 1999; 47: 532–538. 17 Kruizenga HM, Van Tulder MW, Seidell JC, Thijs A, Ader HJ, Van Bokhorst-de van der Schueren MA. Effectiveness and cost-effectiveness of early screening and treatment of malnourished patients. Am J Clin Nutr 2005; 82: 1082–1089. 18 Milte RK, Ratcliffe J, Miller MD, Crotty M. Economic evaluation for protein and energy supplementation in adults: opportunities to strengthen the evidence. Eur J Clin Nutr 2013; 67: 1243–1250. 19 Freijer K, Bours MJ, Nuijten MJ, Poley MJ, Meijers JM, Halfens RJ et al. The economic value of enteral medical nutrition in the management of diseaserelated malnutrition: a systematic review. J Am Med Dir Assoc 2014; 15: 17–29. 20 Norman K, Stobaus N, Gonzalez MC, Schulzke JD, Pirlich M. Hand grip strength: outcome predictor and marker of nutritional status. Clin Nutr 2011; 30: 135–142. 21 Guerra RS, Fonseca I, Pichel F, Restivo MT, Amaral TF. Handgrip strength and associated factors in hospitalized patients. JPEN J Parenter Enteral Nutr 2013 in press. 22 Flood A, Chung A, Parker H, Kearns V, O'Sullivan TA. The use of hand grip strength as a predictor of nutrition status in hospital pat ients. Clin Nutr 2014; 33: 106–114. 23 Mendes J, Azevedo A, Amaral TF. Handgrip strength at admission and time to discharge in medical and surgical inpatients. JPEN J Parenter Enteral Nutr 2014; 38: 481–488. 24 Kilgour RD, Vigano A, Trutschnigg B, Lucar E, Borod M, Morais JA. Handgrip strength predicts survival and is associated with markers of clinical and functional outcomes in advanced cancer patients. Support Care Cancer 2013; 21: 3261–3270. 25 Mahalakshmi VN, Ananthakrishnan N, Kate V, Sahai A, Trakroo M. Handgrip strength and endurance as a predictor of postoperative morbidity in surgical patients: can it serve as a simple bedside test? Int Surg 2004; 89: 115–121. 26 Bohannon RW, Schaubert KL. Test-retest reliability of grip-strength measures obtained over a 12-week interval from community-dwelling elders. J Hand Ther 2005; 18: 426–427. quiz 428.
European Journal of Clinical Nutrition (2015) 187 – 192
27 ASPEN Board of Directors and the Clinical Guidelines Task Force. Guidelines for the use of parenteral and enteral nutrition in adult and pediatric patients. JPEN J Parenter Enteral Nutr 2002; 26: 1SA–138SA. 28 Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373–383. 29 Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing 1972; 1: 233–238. 30 Brodaty H, Pond D, Kemp NM, Luscombe G, Harding L, Berman K et al. The GPCOG: a new screening test for dementia designed for general practice. J Am Geriatr Soc 2002; 50: 530–534. 31 Bonaiuto S, Rocca WA, Lippi A, Luciani P, Giannandrea E, Cavarzeran F et al. Study on the validity of the Hodkinson Abbreviated Mental Test Score (AMTS) in detecting dementia of elderly subjects in appignano (Macerata province), Italy. Arch Gerontol Geriatr 1992; 15(Suppl 1): 75–85. 32 Ottery FD. Patient-Generated Subjective Global Assessment. In: McCallum PD, Polisena CG (eds). The Clinical Guide to Oncology Nutrition. The American Dietetic Association: Chicago, pp 11–23, 2000. 33 Fess EE Grip Strength. American Society of Hand Therapists: Chicago, 1992. 34 Fess EE, Moran CA. Clinical assessment recommendations. American Society of Hand Therapists: Chicago, 1981. 35 Vaz M, Thangam S, Prabhu A, Shetty PS. Maximal voluntary contraction as a functional indicator of adult chronic undernutrition. Br J Nutr 1996; 76: 9–15. 36 Marfell-Jones M, Olds T, Stewart A, Carter L International standards for anthropometric assessment. International Society for the Advancement of Kinanthropometry: Potchefstroom, South Africa, 2006. 37 Guerra RS, Fonseca I, Pichel F, Restivo MT, Amaral TF. Hand length as an alternative measurement of height. Eur J Clin Nutr 2013; 68: 229–233. 38 Kwok T, Whitelaw MN. The use of armspan in nutritional assessment of the elderly. J Am Geriatr Soc 1991; 39: 492–496. 39 Powell-Tuck J, Hennessy EM. A comparison of mid upper arm circumference, body mass index and weight loss as indices of undernutrition in acutely hospitalized patients. Clin Nutr 2003; 22: 307–312. 40 Lee RD, Nieman DC. Assessment of the hospitalized patient. In: Lee RD, Nieman DC (eds). Nutritional Assessment, 4th edn. Mosby: Chicago, IL, USA, pp 226–231, 2007. 41 Pederson D, Gore C. Anthropometry Measurement Error. NSW, University of New South Wales Press: Sydney, Australia, 1996. 42 Ministério da Saúde. Portaria n.º 839-A/2009. In. Diário da República: Portugal, 2009. 43 Ministério da Saúde. Portaria n.º 163/2013. In. Diário da República: Portugal, 2013. 44 Figueiredo F, Dickson ER, Pasha T, Kasparova P, Therneau T, Malinchoc M et al. Impact of nutritional status on outcomes after liver transplantation. Transplantation 2000; 70: 1347–1352. 45 Mendes J, Alves P, Amaral TF. Comparison of nutritional status assessment parameters in predicting length of hospital stay in cancer patients. Clin Nutr 2014; 33: 466–470. 46 Norman K, Stobaus N, Reiss J, Schulzke J, Valentini L, Pirlich M. Effect of sexual dimorphism on muscle strength in cachexia. J Cachexia Sarcopenia Muscle 2012; 3: 111–116. 47 Klek S, Szybinski P, Sierzega M, Szczepanek K, Sumlet M, Kupiec M et al. Commercial enteral formulas and nutrition support teams improve the outcome of home enteral tube feeding. JPEN J Parenter Enteral Nutr 2011; 35: 380–385. 48 Freijer K, Tan SS, Koopmanschap MA, Meijers JM, Halfens RJ, Nuijten MJ. The economic costs of disease related malnutrition. Clin Nutr 2013; 32: 136–141. 49 Thompson JD. DRG prepayment: its purpose and performance. Bull N Y Acad Med 1988; 64: 28–51.
© 2015 Macmillan Publishers Limited
ORIGINAL ARTICLE
Handgrip strength measurement as a predictor of hospitalization costs RS Guerra1,2,3, TF Amaral2,4, AS Sousa3,4, F Pichel3, MT Restivo2, S Ferreira5 and I Fonseca3 BACKGROUND: Undernutrition status at hospital admission is related to increased hospital costs. Handgrip strength (HGS) is an indicator of undernutrition, but the ability of HGS to predict hospitalization costs has yet to be studied. OBJECTIVE: To explore whether HGS measurement at hospital admission can predict patient's hospitalization costs. SUBJECTS/METHODS: A prospective study was conducted in a university hospital. Inpatient's (n = 637) HGS and undernutrition status by Patient-Generated Subjective Global Assessment were ascertained. Multivariable linear regression analysis, computing HGS quartiles by sex (reference: fourth quartile, highest), was conducted in order to identify the independent predictors of hospitalization costs. Costs were evaluated through percentage deviation from the mean cost, after adjustment for patients' characteristics, disease severity and undernutrition status. RESULTS: Being in the first or second HGS quartiles at hospital admission increased patient's hospitalization costs, respectively, by 17.5% (95% confidence interval: 2.7–32.3) and 21.4% (7.5–35.3), which translated into an increase from €375 (58–692) to €458 (161–756). After the additional adjustment for undernutrition status, being in the first or second HGS quartiles had, respectively, an economic impact of 16.6% (1.9–31.2) and 20.0% (6.2–33.8), corresponding to an increase in hospitalization expenditure from €356 (41–668) to €428 (133–724). CONCLUSIONS: Low HGS at hospital admission is associated with increased hospitalization costs of between 16.6 and 20.0% after controlling for possible confounders, including undernutrition status. HGS is an inexpensive, noninvasive and easy-to-use method that has clinical potential to predict hospitalization costs. European Journal of Clinical Nutrition (2015) 69, 187–192; doi:10.1038/ejcn.2014.242; published online 5 November 2014
INTRODUCTION Although this problem has been documented for a long time, a large number of patients are still undernourished when admitted to hospital.1,2 Moreover, nutritional status of a high proportion of patients worsens during hospital stay.3 There is also a considerable amount of scientific evidence showing that undernutrition, often under-recognized and untreated,1,4 is associated with higher hospitalization costs.5–8 Data from multivariable analysis from studies conducted in Portugal (2007) and in Singapore (2012) reveal that the cost of a patient at risk of undernutrition or undernourishment, respectively, is approximately 19.3–24% higher than for a patient not presenting any of these conditions with the same diagnosis-related group (DRG) code.5,6 After controlling for underlying conditions and treatments, data from 2003 to 2004 show that undernutrition was responsible for an increase in AU $1745 per hospital admission.7 It has been recently documented in Croatia that adult undernutrition costs over €97 million annually, accounting for 3.4% of the total Croatian national health-care budget.8 The higher hospitalization costs associated with undernutrition can be attributed to increased risk for infections 9,10 and pressure ulcers,10,11 higher complications rate12 and also longer hospital stay13,14 and the readmission rate.6 Moreover, undernutrition is known to diminish effectiveness of medical treatment15 and to contribute to delayed functional recovery during hospitalization.16
This information on the economic impact of undernutrition emphasizes the importance of identifying and treating undernourishment, as it has been shown that screening and nutritional treatment of undernourished patients can lower hospitalization costs.17–19 There is an increasing evidence that handgrip strength (HGS) is a reliable and attractive method used in clinical daily practice to identify undernutrition.20–22 Undernutrition is associated with impaired hand muscle function and HGS reflects early nutrition deprivation and repletion, before changes in body composition parameters can be detected.20 Moreover, HGS is the most frequently used indicator of muscle function for clinical purposes;20 it has been shown to be a reliable indicator of functional ability20 and also an outcome predictor.20,23–25 Dynamometers used to measure HGS are inexpensive and portable. In addition, the measurement of HGS is noninvasive, easy and quick to perform with no need of specialized professionals, reliable 26 and shows low intra- and inter-observers variability.26 Because of the budgetary constraints that health-care systems face today,2 hospitalization undernutrition-related costs are a key issue in health-care management. Therefore, it is of the utmost importance to seek a simple and reliable tool to predict hospitalization expenditure so that a timely and effective health-care plan can be implemented. This study was designed to explore the ability of HGS measurement to predict patient's hospitalization costs,
1 Departamento de Bioquímica, Faculdade de Medicina da Universidade do Porto, Porto, Portugal; 2UISPA-IDMEC, Faculdade de Engenharia da Universidade do Porto, Porto, Portugal; 3Unidade de Nutrição, Centro Hospitalar do Porto, Porto, Portugal; 4Faculdade de Ciências da Nutrição e Alimentação da Universidade do Porto, Porto, Portugal and 5 Serviço de Informação de Gestão, Centro Hospitalar do Porto, Porto, Portugal. Correspondence: RS Guerra, Departamento de Bioquímica, Faculdade de Medicina da Universidade do Porto, Al. Professor Hernâni, Porto 4200-319, Portugal. E-mail: [email protected] Received 02 June 2014; revised 16 September 2014; accepted 27 September 2014; published online 5 November 2014
Handgrip strength and hospitalization costs RS Guerra et al
188 considering the possible confounder effect of patients' characteristics, disease severity and undernutrition status. SUBJECTS AND METHODS Study population and design A prospective study was conducted in a Portuguese university hospital between July 2011 and June 2013. A consecutive sampling approach was used. From the daily list of patients admitted to each ward, those who met the inclusion criteria were invited to participate in the study. Patients were eligible to participate in the study if they were ⩾ 18 years old, Caucasian, with an expected length of stay 424 h, conscious, cooperative and able to provide written informed consent. Patients unable to perform the HGS technique, defined as unable to understand verbal instructions or having a condition in which the patient could not perform the technique correctly and also patients with critical illness, defined as failure of at least one vital organ27 and admitted to intensive care units, were excluded from the study. Pregnant women, individuals in isolation, those who were admitted for procedures that could put them in a critical situation and those with hemodynamic instability at the time of evaluation were also excluded. Therefore, inpatients from angiology and vascular surgery, cardiology, digestive, non-digestive and hepato-biliary surgeries, endocrinology, gastroenterology, internal medicine, nephrology, orthopedics, otolaryngology and urology wards were considered eligible for this study. Sample size for linear regression was calculated assuming an effect size of 0.15, a statistical power of 80% and a level of significance equal to 0.05. The minimum required sample size was 113 individuals. In order to have a representative sample of the hospitalized patients who met the inclusion criteria, data collection took place until the number of patients corresponded to twice the total number of beds of all the selected wards (n = 711). Therefore, this study sample had sufficient statistical power to identify the independent variables associated with hospitalization costs. Inpatients transferred from intensive care units to the hospital ward where data were collected and inpatients transferred from surgical to medical wards (n = 48) were not included in data analysis, given that these participants' hospitalization percentage of cost deviation was above the superior limit of 95% confidence interval (CI) of cost estimates. Given the likelihood of providing inaccurate information, patients with possible cognitive impairment (n = 13) and also patients with missing data (n = 13) were excluded from data analysis. The final study sample was composed of 637 participants.
Ethics This research was conducted according to the guidelines laid down in the Declaration of Helsinki and approved by the Institutional Review Board and the Ethics Committee of Centro Hospitalar do Porto. Written informed consent was obtained from all study participants.
Data collection Demographical data and clinical history were obtained by consulting the patient's clinical file. DRG codes were obtained from hospital records. The remaining data were collected within 72 h of admission to the hospital by two previously trained nutritionists using a structured questionnaire. The education level was evaluated by the number of completed schooling years and the following classes were created: 0–4, 5–12 and 412 school years. Marital status was categorized as single, married or in a common-law marriage, divorced and widowed. Disease severity was evaluated using the Charlson comorbidity index.28 Medical discharge diagnoses in the patient's clinical record were obtained by a previously trained interviewer and used to classify comorbidity conditions. The Charlson index takes into account the number and the seriousness of comorbid diseases, each scored from 0 to 6.28 Possibility of cognitive impairment was assessed using the Abbreviated Mental Test.29 This test consists of 10 questions, each scored one point if correct, with a cutoff score of 7 or 8 out of 10 discriminating between cognitive impairment and normality in older adults.29 Although the Abbreviated Mental Test was developed to discriminate cognitive impairment of older adults, it has also been applied in adults younger than 65 years.30 In the present study, a score o6 was considered denotative of cognitive impairment because it has shown the best combination of sensitivity and specificity in a mixed sample of adults and older adults.31 Participants' undernutrition status was evaluated by Patient-Generated Subjective Global Assessment.32 European Journal of Clinical Nutrition (2015) 187 – 192
Nondominant HGS was measured in kgf with a calibrated Jamar Hand Dynamometer (Sammons Preston, Bolingbrook, IL, USA), recommended by the American Society of Hand Therapists.33 Measurements were performed in a chair or on a bed, with the shoulder adducted and neutrally rotated, elbow flexed to an angle of 90º and the forearm and wrist in neutral position.34 Each participant performed three measurements with a 1 min pause between measurements and the maximum value was chosen as the HGS value.35 When the individual was unable to perform the measurement with the nondominant hand, the dominant hand was used (n = 53). Standing height (cm) was obtained with a metal tape (Rosscraft Innovations Incorporated, Surrey, Canada), with 0.1-cm resolution and a headboard.36 For the participants with visible kyphosis or when it was impossible to measure standing height (n = 312), recumbent height was obtained.37,38 Body weight (kg)36 was measured with a calibrated portable beam scale with 0.5-kg resolution. For participants on dialytic therapies (n = 19), dry body weight was used. When it was not possible to weigh a patient (n = 148), body weight was estimated from height and mid-upper arm circumference.39 Mid-upper arm circumference was measured with a metal tape (Rosscraft Innovations Incorporated, Surrey, Canada), with 0.1-cm resolution.40 The intra- and inter-observer technical error of measurement was obtained for all anthropometric measurements, respectively, in 17 and 18 individuals. Intra-observer error varied between 0.2 and 0.6%, and interobserver error varied between 0 and 1.4%, respectively, values considered acceptable for trained anthropometrists.41
Statistical analysis The Kolmogorov–Smirnov test was used to evaluate the normality of variable distribution. Continuous normally distributed variables were described as mean and s.d. and non-normal as median and interquartile range. Categorical variables were reported as frequencies. HGS was summarized using quartiles by sex, according to the cutoffs of sample distribution stratified by sex. Cutoffs were 12.0, 16.3 and 21.9 kgf for women and 25.8, 32.0 and 37.8 kgf for men. HGS varied between 1.0 and 35.1 kgf for women; median (interquartile range) was equal to 16.4 (9.6) kgf. For men, HGS values varied between 1.0 and 61.0 kgf, with a median (interquartile range) of 32.0 (12.0) kgf. Hospitalization cost was calculated for each inpatient on the basis of discharge DRG codes. The DRG system is used to calculate hospital reimbursements, with the amounts determined on the basis of a relative weight value. This weight value reflects the main diagnostic, surgical interventions, associated pathologies and complications, clinical procedures, medium length of hospital stay, patient age and sex and discharge destination. The information about DRG codes and its amounts was obtained from Portuguese Ministerial Directive number 839-A, 31 July 2009,42 used for data obtained between 2011 and 2012 and number 163, 24 April 2013, 43 which replaced the latter directive, was used for data obtained in 2013. The percentage of cost deviation was calculated from the difference between the cost of each inpatient and the cost of a patient with relative weight of one (€2396 for data obtained between 2011 and 2012; €2142 for data obtained in 2013). Percentage of cost deviation was summarized using quartiles according to the cutoffs of the sample distribution (−46.0, 1.3 and 49.0%). In order to identify variables associated with HGS and with the percentage of cost deviation, patients' demographic and clinical characteristics were compared across HGS quartiles and across the percentage of cost deviation quartiles using one-way Anova (or the Kruskal–Wallis test) for continuous variables and the Pearson χ2-test for categorical variables. Multivariable linear regression analysis using the stepwise method was performed to identify the independent variables of percentage of cost deviation. The following variables were included in the model: HGS quartiles, age, education, professional activity, marital status, hospitalization in medical or surgical wards, Abbreviated Mental Test and Charlson comorbidity index scores. As HGS is an indicator of the nutrition status,1,20–22 a multivariable analysis was conducted including all the mentioned variables and adjusting for undernutrition status. The interaction between hospitalization in medical or surgical wards and HGS quartiles was also tested. Results were considered significant when Po0.05. Statistical analyses were conducted using the Software Package for Social Sciences (SPSS) for Windows, version 21.0; SPSS, Inc, an IBM Company, Chicago, IL, USA. © 2015 Macmillan Publishers Limited
Handgrip strength and hospitalization costs RS Guerra et al
RESULTS The characteristics of the 637 patients (aged between 18 and 91 years old) stratified by HGS quartiles are presented in Table 1. Participants with lower HGS values were older, presented lower Abbreviated Mental Test and higher Charlson comorbidity index scores. A lower percentage of participants in work and higher proportions of widows, patients with ⩽ 4 years of schooling and admitted in medical wards, were observed in the lower quartiles of HGS compared with the higher quartiles of HGS. Moreover, a higher proportion of patients who were severely undernourished was observed in the three lower quartiles of HGS. In the two lower HGS quartiles, a positive percentage of cost deviation was found. Thus, participants with low HGS have a mean cost higher compared with the cost of a patient with relative weight of one. In contrast, the percentage of cost deviation was negative in the two higher quartiles of HGS, which means that participants with high HGS have a mean cost lower compared with the cost of a patient with relative weight of one. Patient's hospitalization costs within this sample varied between €393 and €20447, with a median (interquartile range) equal to €2428 (€3030). Participant's characteristics according to the percentage of cost deviation quartiles are presented in Table 2. Participants in the two lower quartiles of percentage of cost deviation were younger, single or divorced, had a professional activity and a higher number of schooling years. Otherwise, Table 1.
189 a higher proportion of inpatients admitted in medical wards was observed in the lower quartiles of percentage of cost deviation. Patients in the two lower quartiles of percentage of cost deviation presented better nutrition status by Patient-Generated Subjective Global Assessment, compared with the two higher quartiles. A higher proportion of patients with high HGS in the lower quartiles of percentage of cost deviation was also observed. The multivariable linear regression models computed to predict the percentage of cost deviation are presented in Table 3. In model 1, the predictors of higher hospitalization costs were advanced age, hospitalization in surgical wards and being in the first (the lowest) or in the second HGS quartiles. In model 2, it is shown that advanced age, hospitalization in surgical wards, being in the first (the lowest) or in the second HGS quartiles and severe undernutrition predicted higher percentage of cost deviation. The interaction between hospitalization in medical or surgical wards and HGS quartiles was tested, but no significant effect was found in any of the fitted models (P ⩾ 0.349). According to our estimation, and after controlling for possible confounders, being in the first or second HGS quartiles at hospital admission is related to increased odds of having higher hospitalization costs of between €375 (95% CI: €58–692) and €458 (95% CI: €161–756). After the additional adjustment for undernutrition status, being in the first or second HGS quartiles
Sample characteristics according to handgrip strength quartiles Handgrip strength quartilesa (n = 637)
Characteristic
1st (lowest) (n = 154)
2nd (n = 166)
3rd (n = 151)
4th (highest) (n = 166)
P
Sex, n (%) Women Men
72 (46.8) 82 (53.2)
72 (43.4) 94 (56.6)
73 (48.3) 78 (51.7)
73 (44.0) 93 (56.0)
Age (years), median (IQR)
65 (21)
63 (19)
57 (21)
47 (24)
o 0.001c
Having a professional activity, n (%)
25 (16.2)
35 (21.1)
49 (32.5)
82 (49.4)
o 0.001b
Education (years), n (%) 0–4 5–12 412
86 (56.2) 59 (38.6) 8 (5.2)
74 (44.8) 76 (46.1) 15 (9.1)
60 (40.0) 76 (50.7) 14 (9.3)
50 (30.1) 98 (59.0) 18 (10.8)
0.001b
Marital status, n (%) Single Married/common-law marriage Divorced Widow
23 99 12 20
Hospital wards, n (%) Medical Surgical
90 (33.1) 64 (17.5)
(14.9) (64.3) (7.8) (13.0)
22 107 13 24
(13.3) (64.5) (7.8) (14.5)
67 (24.6) 99 (27.1)
26 104 13 8
(17.2) (68.9) (8.6) (5.3)
64 (23.5) 87 (23.8)
0.790b
(19.9) (65.1) (10.8) (4.2)
0.032b
51 (18.8) 115 (31.5)
o 0.001b
33 108 18 7
Abbreviated Mental Test (score), median (IQR)
9 (2)
9 (1)
10 (1)
10 (1)
0.006c
Charlson comorbidity index (score), median (IQR)
2 (2)
2 (3)
2 (3)
1 (2)
o 0.001c
126 (75.9) 17 (10.2) 23 (13.9)
o 0.001b
PG-SGA, n (%) Not undernourished Moderate or suspected undernutrition Severe undernutrition
55 (37.7) 58 (35.7) 41 (26.6)
73 (44.0) 48 (28.9) 45 (27.1)
83 (55.0) 34 (22.5) 34 (22.5)
Body mass index, kg/m2, mean (s.d.)
25.7 (5.1)
26.6 (5.5)
26.0 (5.0)
27.2 (5.3)
Percentage of cost deviation, €, median (IQR)
23.6 (94.2)
31.9 (94.5)
− 10.8 (94.1)
− 12.2 (101.2)
0.052d o 0.001c
Abbreviations: IQR, interquartile range; PG-SGA, Patient-Generated Subjective Global Assessment. aCutoffs: 12.0, 16.3 and 21.9 kgf for women; 25.8, 32.0 and 37.8 kgf for men. bPearson χ2-test. cKruskal–Wallis test. dOne-way ANOVA.
© 2015 Macmillan Publishers Limited
European Journal of Clinical Nutrition (2015) 187 – 192
Handgrip strength and hospitalization costs RS Guerra et al
190 Table 2.
Sample characteristics according to percentage of cost deviation quartiles Percentage of cost deviation quartiles (n = 637)
Characteristic
1st −509.1, −46.0 (n = 159)
2nd −45.9, 1.3 (n = 163)
3rd 1.4, 49.0 (n = 156)
4th 49.1, 88.3 (n = 159)
P
Sex, n (%) Women Men
76 (47.8) 83 (52.2)
73 (44.8) 90 (55.2)
64 (41.0) 92 (59.0)
77 (48.4) 82 (51.6)
Age (years), median (IQR)
53 (29)
52 (23)
63 (19)
62 (16)
o 0.001b
Having a professional activity, n (%)
61 (38.4)
64 (39.3)
33 (21.2)
33 (20.8)
o 0.001a
Education (years), n (%) 0–4 5–12 412
52 (32.9) 87 (55.1) 19 (12.0)
62 (38.0) 86 (52.8) 15 (9.2)
74 (47.7) 67 (43.2) 14 (9.0)
82 (51.9) 69 (43.7) 7 (4.4)
0.008a
Marital status, n (%) Single Married/common-law marriage Divorced Widow
33 102 16 8
(20.8) (64.2) (10.1) (5.0)
34 100 17 12
(20.9) (61.3) (10.4) (7.4)
23 103 12 18
(14.7) (66.0) (7.7) (11.5)
0.536a
(8.8) (71.1) (6.9) (13.2)
0.019a
41 (25.8) 118 (74.2)
o 0.001a
14 113 11 21
Hospital wards, n (%) Medical Surgical
86 (54.1) 73 (45.9)
77 (47.2) 86 (52.8)
68 (43.6) 88 (56.4)
Abbreviated mental test (score), median (IQR)
10 (1)
10 (1)
10 (1)
9 (1)
0.526b
2 (2)
2 (2)
2 (2)
2 (3)
0.080b
Charlson comorbidity index (score), median (IQR) PG-SGA, n (%) Not undernourished Moderate or suspected undernutrition Severe undernutrition
102 (64.2) 32 (20.1) 25 (15.7)
Body mass index, kg/m2, mean (s.d.)
26.3 (5.3)
26.0 (5.2)
27.0 (5.5)
26.4 (5.1)
0.429c
Handgrip strength, kgf, median (IQR)
25.0 (17.0)
24.0 (15.9)
24.0 (15.8)
22.0 (16.6)
0.072b
Handgrip strength quartilesd, kgf, n (%) 1st (lowest) 2nd 3rd 4th (highest)
34 35 38 52
(21.4) (22.0) (23.9) (32.7)
93 (57.1) 46 (28.2) 24 (14.7)
35 33 48 47
(21.5) (20.2) (29.4) (28.8)
73 (46.8) 40 (25.8) 43 (27.6)
41 45 32 38
(26.3) (28.8) (20.5) (24.4)
72 (45.3) 36 (22.6) 51 (32.1)
44 53 33 29
(27.7) (33.3) (20.8) (18.2)
o 0.001a
0.021a
Abbreviations: IQR, interquartile range; PG-SGA, Patient-Generated Subjective Global Assessment. aPearson χ2-test. bKruskal–Wallis test. cOne-way ANOVA. d Cutoffs: 12.0, 16.3 and 21.9 kgf for women; 25.8, 32.0 and 37.8 kgf for men.
had, respectively, an economic impact of between €356 (95% CI: €41–668) and €428 (95% CI: €133–724). DISCUSSION Findings from the present study clearly demonstrate that patients with lower HGS at hospital admission have higher hospitalization costs. This is, as far as we are concerned, the first study to show that HGS at hospital admission predicts deviations from the hospitalization cost of a standard patient with relative weight of one. HGS is considered as an indicator of functional and nutrition status.1,20–22 It is therefore important to highlight that low HGS was a significant predictor of higher hospitalization costs, even after controlling for undernutrition status. It was also shown in the model not adjusted for undernutrition status that having low HGS, advanced age and being hospitalized in surgical wards independently predicted higher hospitalization costs. Moreover, an updated European Journal of Clinical Nutrition (2015) 187 – 192
estimate of hospitalization costs associated with patient's severe undernutrition according to Patient-Generated Subjective Global Assessment is provided, highlighting the economic burden of undernutrition at hospital setting. Examining the explained variation assessed by the adjusted coefficients of multiple determination (adjusted R2), it is also noteworthy that the two models share a similar explanatory power in predicting hospital costs. Undernutrition, after the adjustment for several confounders, represented an independent increase in costs of 19.3% in Portugal in 2007,5 whereas in Singapore the average cost of hospitalization was 24% higher for undernourished patients in 2012.6 Compared with these previous studies, low HGS in the present sample represented a similar independent increase in hospitalization costs between 16.6 and 21.4%. Exclusion criteria should be acknowledged because the sample restriction to patients able to provide written informed consent limits the generalization of study results for hospitalized patients who are not in this situation. © 2015 Macmillan Publishers Limited
Handgrip strength and hospitalization costs RS Guerra et al
191 Table 3. Multivariable linear regression models for prediction of hospitalization cost deviationa,b Regression coefficient Model 1d Age (years) Hospital wards (surgical vs medical)
0.80 27.3
95% CI
0.43–1.17 o0.001 15.8–38.8 o0.001
Handgrip strength quartilese (reference: 4th quartile) 1st (lowest) 17.5 2.7–32.3 2nd 21.4 7.5–35.3 3rd 0.07 − Model 2f PG-SGA (reference: not undernourished) Moderate/suspected undernutrition Severe undernutrition Age (years) Hospital wards (surgical vs medical)
0.06 24.7 0.70 24.4
Pc
−
0.020 0.003 0.159
0.151
11.3–38.2 o0.001 0.33–1.07 o0.001 16.1–38.8 o0.001
Handgrip strength quartilese (reference: 4th quartile) 1st (lowest) 16.6 1.9–31.2 2nd 20.0 6.2–33.8 3rd 0.06 −
0.027 0.005 0.209
Abbreviations: CI, confidence interval; PG-SGA, Patient-Generated Subjective Global Assessment. aCalculated as a percentage from the difference between the cost of each inpatient and the cost of a patient with relative weight of one. bVariables included the following: Model 1, HGS quartiles (fourth HGS quartile, highest, was used as reference), age, education (having 412 schooling years was used as reference), professional activity (having a professional activity was used as reference), marital status (being married or in common-law marriage was used as reference), hospitalization in medical or surgical wards (hospitalization in medical wards was used as reference), the Abbreviated Mental Test score, the Charlson comorbidity index score. Model 2, all the aforementioned variables and undernutrition status by PG-SGA (‘not undernourished’ was used as reference). cP-value by linear regression analysis. dAdjusted R2 = 0.081. eCutoffs: 12.0, 16.3 and 21.9 kgf for women; 25.8, 32.0 and 37.8 kgf for men. fAdjusted R2 = 0.099.
Despite the higher costs that undernutrition implies, identification of undernourished patients and proper nutritional treatment can reduce hospitalization costs.17–19 A study from 2005 showed that the incremental cost of 1 less day in hospital for undernourished patients due to nutritional treatment was €76, whereas the mean cost of 1 day in hospital was €476 for university hospitals and €337 for peripheral hospitals.17 This study stated that the incremental costs of undernutrition screening and treatment to reduce the length of hospital stay by 1 day were higher in undernourished patients who had HGS higher than the standard compared with those of undernourished patients with HGS lower than the standard.17 This is probably because frail patients, that is, patients simultaneously undernourished and with low HGS would benefit more from nutritional intervention.17 In a study conducted among patients subjected to liver transplant, the assessment of preoperative nutrition status included HGS, but no single or combination of parameters predicted global resource utilization, which is an index of cost.44 Patients identified as undernourished were given nutritional support, which can negate the adverse effects of undernutrition on post-transplant outcomes.44 Other explanations for this finding are that only five individuals (9.4% of study sample) were severely undernourished and liver transplant is a high-cost situation where HGS could have a lower predictive value.44 Some strengths of the present study can be enumerated. The association between HGS and hospitalization cost was adjusted for potential confounders, including undernutrition status. Also, © 2015 Macmillan Publishers Limited
the study sample is composed of a large number of inpatients within a varying age range (18–91 years old) from a variety of hospital wards and therefore ensuring a wide spectrum of diagnoses and relevant pathologies. Moreover, present sample HGS values and quartile cutoffs are comparable to other samples of inpatients.45,46 These aspects strengthen the external validity of the present study results for other hospitalized patients. The DRG system was shown to underestimate the real hospitalization costs.47 Moreover, only direct hospitalization costs are considered, whereas indirect costs, such as productivity loss due to time off work or societal costs,48 are not. Nevertheless, this methodology has been used extensively to estimate undernutrition costs, 5,6,8 as it allows to cluster patients with a variety of diagnoses and procedures into diagnostic groups. Cases within a group will have a similar level of complexity, and their treatment is expected to have similar costs.49 For these reasons, the DRG cost estimation methodology was used in this study. Until now, the ability of HGS measurement at admission to predict patient's hospitalization cost has never been assessed, justifying the current study. Results first show that low HGS is associated with higher hospitalization costs, considering the possible confounder effect of other factors such as hospitalization in surgical wards and undernutrition. Patients with a low HGS at hospital admission have an independent increase in hospitalization costs ranging from 16.6% (95% CI: 1.9–31.2%) to 21.4% (95% CI: 7.5–35.3%), representing €356 (95% CI: €41–668) and €458 (95% CI: €161–756). The early identification of patients with low HGS offers the opportunity to treat them, namely by nutrition intervention, which can improve their clinical progress and outcome and possibly reduce hospitalization costs. Future studies should compare HGS ability in predicting costs with that of other undernutrition measures and tools. Also, experimental studies should evaluate the cost-effectiveness of nutritional treatment of patients who are identified with low HGS values at hospital admission. HGS is an inexpensive, noninvasive and easy-to-use method that has clinical potential to predict hospitalization costs. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank Centro Hospitalar do Porto and all ward directors for facilitating the data collection. Rita S Guerra as a PhD student received a scholarship from FCT—Fundação para a Ciência e a Tecnologia under the project (SFRH/BD/61656/2009).
REFERENCES 1 White JV, Guenter P, Jensen G, Malone A, Schofield M. Consensus statement: Academy of Nutrition and Dietetics and American Society for Parenteral and Enteral Nutrition: characteristics recommended for the identification and documentation of adult malnutrition (undernutrition). JPEN J Parenter Enteral Nutr 2012; 36: 275–283. 2 Stratton RJ, Elia M. Encouraging appropriate, evidence-based use of oral nutritional supplements. Proc Nutr Soc 2010; 69: 477–487. 3 Rasmussen HH, Holst M, Kondrup J. Measuring nutritional risk in hospitals. Clin Epidemiol 2010; 2: 209–216. 4 Tappenden KA, Quatrara B, Parkhurst ML, Malone AM, Fanjiang G, Ziegler TR. Critical role of nutrition in improving quality of care: an interdisciplinary call to action to address adult hospital malnutrition. J Acad Nutr Diet 2013; 113: 1219–1237. 5 Amaral TF, Matos LC, Tavares MM, Subtil A, Martins R, Nazare M et al. The economic impact of disease-related malnutrition at hospital admission. Clin Nutr 2007; 26: 778–784. 6 Lim SL, Ong KC, Chan YH, Loke WC, Ferguson M, Daniels L. Malnutrition and its impact on cost of hospitalization, length of stay, readmission and 3-year mortality. Clin Nutr 2012; 31: 345–350.
European Journal of Clinical Nutrition (2015) 187 – 192
Handgrip strength and hospitalization costs RS Guerra et al
192 7 Rowell DS, Jackson TJ. Additional costs of inpatient malnutrition, Victoria, Australia, 2003-2004. Eur J Health Econ 2011; 12: 353–361. 8 Benkovic V, Kolcic I, Ivicevic Uhernik A, Vranesic Bender D, Oreb I, Stevanovic R et al. The economic burden of disease-related undernutrition in selected chronic diseases. Clin Nutr 2014; 33: 689–693. 9 Schneider SM, Veyres P, Pivot X, Soummer AM, Jambou P, Filippi J et al. Malnutrition is an independent factor associated with nosocomial infections. Br J Nutr 2004; 92: 105–111. 10 Fry DE, Pine M, Jones BL, Meimban RJ. Patient characteristics and the occurrence of never events. Arch Surg 2010; 145: 148–151. 11 Banks MD, Graves N, Bauer JD, Ash S. The costs arising from pressure ulcers attributable to malnutrition. Clin Nutr 2010; 29: 180–186. 12 Correia MI, Waitzberg DL. The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis. Clin Nutr 2003; 22: 235–239. 13 Leandro-Merhi VA, de Aquino JL, Sales Chagas JF. Nutrition status and risk factors associated with length of hospital stay for surgical patients. JPEN J Parenter Enteral Nutr 2011; 35: 241–248. 14 Pavic T, Ljubicic N, Stojsavljevic S, Krznaric Z. Nutritional screening model in tertiary medical unit in Croatia. Ann Nutr Metab 2012; 61: 65–69. 15 Jensen GL, Mirtallo J, Compher C, Dhaliwal R, Forbes A, Grijalba RF et al. Adult starvation and disease-related malnutrition: a proposal for etiologybased diagnosis in the clinical practice setting from the International Consensus Guideline Committee. JPEN J Parenter Enteral Nutr 2010; 34: 156–159. 16 Covinsky KE, Martin GE, Beyth RJ, Justice AC, Sehgal AR, Landefeld CS. The relationship between clinical assessments of nutritional status and adverse outcomes in older hospitalized medical patients. J Am Geriatr Soc 1999; 47: 532–538. 17 Kruizenga HM, Van Tulder MW, Seidell JC, Thijs A, Ader HJ, Van Bokhorst-de van der Schueren MA. Effectiveness and cost-effectiveness of early screening and treatment of malnourished patients. Am J Clin Nutr 2005; 82: 1082–1089. 18 Milte RK, Ratcliffe J, Miller MD, Crotty M. Economic evaluation for protein and energy supplementation in adults: opportunities to strengthen the evidence. Eur J Clin Nutr 2013; 67: 1243–1250. 19 Freijer K, Bours MJ, Nuijten MJ, Poley MJ, Meijers JM, Halfens RJ et al. The economic value of enteral medical nutrition in the management of diseaserelated malnutrition: a systematic review. J Am Med Dir Assoc 2014; 15: 17–29. 20 Norman K, Stobaus N, Gonzalez MC, Schulzke JD, Pirlich M. Hand grip strength: outcome predictor and marker of nutritional status. Clin Nutr 2011; 30: 135–142. 21 Guerra RS, Fonseca I, Pichel F, Restivo MT, Amaral TF. Handgrip strength and associated factors in hospitalized patients. JPEN J Parenter Enteral Nutr 2013 in press. 22 Flood A, Chung A, Parker H, Kearns V, O'Sullivan TA. The use of hand grip strength as a predictor of nutrition status in hospital pat ients. Clin Nutr 2014; 33: 106–114. 23 Mendes J, Azevedo A, Amaral TF. Handgrip strength at admission and time to discharge in medical and surgical inpatients. JPEN J Parenter Enteral Nutr 2014; 38: 481–488. 24 Kilgour RD, Vigano A, Trutschnigg B, Lucar E, Borod M, Morais JA. Handgrip strength predicts survival and is associated with markers of clinical and functional outcomes in advanced cancer patients. Support Care Cancer 2013; 21: 3261–3270. 25 Mahalakshmi VN, Ananthakrishnan N, Kate V, Sahai A, Trakroo M. Handgrip strength and endurance as a predictor of postoperative morbidity in surgical patients: can it serve as a simple bedside test? Int Surg 2004; 89: 115–121. 26 Bohannon RW, Schaubert KL. Test-retest reliability of grip-strength measures obtained over a 12-week interval from community-dwelling elders. J Hand Ther 2005; 18: 426–427. quiz 428.
European Journal of Clinical Nutrition (2015) 187 – 192
27 ASPEN Board of Directors and the Clinical Guidelines Task Force. Guidelines for the use of parenteral and enteral nutrition in adult and pediatric patients. JPEN J Parenter Enteral Nutr 2002; 26: 1SA–138SA. 28 Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373–383. 29 Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing 1972; 1: 233–238. 30 Brodaty H, Pond D, Kemp NM, Luscombe G, Harding L, Berman K et al. The GPCOG: a new screening test for dementia designed for general practice. J Am Geriatr Soc 2002; 50: 530–534. 31 Bonaiuto S, Rocca WA, Lippi A, Luciani P, Giannandrea E, Cavarzeran F et al. Study on the validity of the Hodkinson Abbreviated Mental Test Score (AMTS) in detecting dementia of elderly subjects in appignano (Macerata province), Italy. Arch Gerontol Geriatr 1992; 15(Suppl 1): 75–85. 32 Ottery FD. Patient-Generated Subjective Global Assessment. In: McCallum PD, Polisena CG (eds). The Clinical Guide to Oncology Nutrition. The American Dietetic Association: Chicago, pp 11–23, 2000. 33 Fess EE Grip Strength. American Society of Hand Therapists: Chicago, 1992. 34 Fess EE, Moran CA. Clinical assessment recommendations. American Society of Hand Therapists: Chicago, 1981. 35 Vaz M, Thangam S, Prabhu A, Shetty PS. Maximal voluntary contraction as a functional indicator of adult chronic undernutrition. Br J Nutr 1996; 76: 9–15. 36 Marfell-Jones M, Olds T, Stewart A, Carter L International standards for anthropometric assessment. International Society for the Advancement of Kinanthropometry: Potchefstroom, South Africa, 2006. 37 Guerra RS, Fonseca I, Pichel F, Restivo MT, Amaral TF. Hand length as an alternative measurement of height. Eur J Clin Nutr 2013; 68: 229–233. 38 Kwok T, Whitelaw MN. The use of armspan in nutritional assessment of the elderly. J Am Geriatr Soc 1991; 39: 492–496. 39 Powell-Tuck J, Hennessy EM. A comparison of mid upper arm circumference, body mass index and weight loss as indices of undernutrition in acutely hospitalized patients. Clin Nutr 2003; 22: 307–312. 40 Lee RD, Nieman DC. Assessment of the hospitalized patient. In: Lee RD, Nieman DC (eds). Nutritional Assessment, 4th edn. Mosby: Chicago, IL, USA, pp 226–231, 2007. 41 Pederson D, Gore C. Anthropometry Measurement Error. NSW, University of New South Wales Press: Sydney, Australia, 1996. 42 Ministério da Saúde. Portaria n.º 839-A/2009. In. Diário da República: Portugal, 2009. 43 Ministério da Saúde. Portaria n.º 163/2013. In. Diário da República: Portugal, 2013. 44 Figueiredo F, Dickson ER, Pasha T, Kasparova P, Therneau T, Malinchoc M et al. Impact of nutritional status on outcomes after liver transplantation. Transplantation 2000; 70: 1347–1352. 45 Mendes J, Alves P, Amaral TF. Comparison of nutritional status assessment parameters in predicting length of hospital stay in cancer patients. Clin Nutr 2014; 33: 466–470. 46 Norman K, Stobaus N, Reiss J, Schulzke J, Valentini L, Pirlich M. Effect of sexual dimorphism on muscle strength in cachexia. J Cachexia Sarcopenia Muscle 2012; 3: 111–116. 47 Klek S, Szybinski P, Sierzega M, Szczepanek K, Sumlet M, Kupiec M et al. Commercial enteral formulas and nutrition support teams improve the outcome of home enteral tube feeding. JPEN J Parenter Enteral Nutr 2011; 35: 380–385. 48 Freijer K, Tan SS, Koopmanschap MA, Meijers JM, Halfens RJ, Nuijten MJ. The economic costs of disease related malnutrition. Clin Nutr 2013; 32: 136–141. 49 Thompson JD. DRG prepayment: its purpose and performance. Bull N Y Acad Med 1988; 64: 28–51.
© 2015 Macmillan Publishers Limited
