S. Giryes, E. Leibovitz, Z. Matas, S. Fridman, D. Gavish, B. Shalev, Z. Ziv-Nir, Y. Berlovitz and M. Boaz
Background: Depending on the definition used, malnutrition is prevalent among 20¨C50% of hospitalized patients. Routine nutritional screening is necessary to identify patients with or at increased risk for malnutrition. The Nutrition Risk Screening (NRS 2002) has been recommended as an efficient tool to identify the risk of malnutrition in adult inpatients.
Objectives: To utilize the NRS 2002 to estimate the prevalence of malnutrition among newly hospitalized adult patients, and to identify risk factors for malnutrition.
Methods: During a 5 week period, all adult patients newly admitted to all inpatient departments (except Maternity and Emergency) at Wolfson Medical Center, Holon, were screened using the NRS 2002. An answer of yes recorded for any of the Step 1 questions triggered the Step 2 screen on which an age-adjusted total score ¡Ý 3 indicated high malnutrition risk.
Results: Data were obtained from 504 newly hospitalized adult patients, of whom 159 (31.5%) were identified as high risk for malnutrition. Malnutrition was more prevalent in internal medicine than surgical departments: 38.6% vs. 19.1% (P < 0.001). Body mass index was within the normal range among subjects at high risk for malnutrition: 23.9 ¡À 5.6 kg/m2 but significantly lower than in subjects at low malnutrition risk: 27.9 ¡À 5.3 kg/m2 (P < 0.001). Malnutrition risk did not differ by gender or smoking status, but subjects at high malnutrition risk were significantly older (73.3 ¡À 16.2 vs. 63.4 ¡À 18.4 years, P < 0.001). Total protein, albumin, total cholesterol, low density lipoprotein-cholesterol, hemoglobin and %lymphocytes were all significantly lower, whereas urea, creatinine and %neutrophils were significantly higher in patients at high malnutrition risk.
Conclusions: Use of the NRS 2002 identified a large proportion of newly hospitalized adults as being at high risk for malnutrition. These findings indicate the need to intervene on a system-wide level during hospitalization.
I. Potasman, G. Naftali and M. Grupper
Background: Overuse and abuse of antibiotics is a major cause of microbial resistance. Within the hospital setting such overuse necessitates real-time supervision by infectious diseases (ID) specialists.
Objectives: To evaluate the impact of a recently introduced computerized antibiotic authorization system on the pharmacy budget.
Methods: The study was performed in a 400 bed university hospital. With the new system, antibiotic requests are entered electronically by the ward physician and reviewed within minutes to hours by ID specialists. The feedbacks are seen in the wards and pharmacy. Successive years, one before and the other after introduction of the system, were compared.
Results: During the first year with the new system 7167 antibiotic requests were entered 20% of them were rejected, mainly for improper indication (43% of the rejections). During that year the antibiotic expenditure was reduced by 17%, compared to the previous year (~equal to 200,000 US$), and was against the trend of the last 5 years. Of the 35 antibiotics under the control of the ID team, the use of 7 was probably curtailed by the supervision. Pareto analysis revealed that four drugs constituted > 50% of the pharmacy’s expenses. The mortality rate (per 1000 hospitalization days) during those 2 years fell from 4.0 to 3.8.
Conclusions: Computerized antibiotic control by ID specialists is a feasible cost-saving new modality that may help reduce unnecessary antibiotic prescriptions.
R. Marom, R. Lubetzky, F.B. Mimouni, H. Bassan, L. Ben Sira, I. Berger, S. Dollberg and D. Mandel
Background: Infants with severe intraventricular-periventricular hemorrhage (IVH) have higher absolute nucleated red blood cell counts (aNRBC) at birth (a marker of intrauterine hypoxia) than controls. Periventricular leukomalacia (PVL) is known to be associated with prenatal and postnatal events. Whether PVL is also linked to intrauterine hypoxia is unknown.
Objectives: To test the hypothesis that infants with PVL have higher aNRBC counts at birth than controls.
Methods: We studied 14 very low birth weight infants with PVL and compared them with 14 pair-matched controls without PVL. Head ultrasound scans were performed in all infants on days 3–5 and 21–25 of life. Paired tests, Fisher exact tests and stepwise logistic regression were performed for analysis.
Results: Groups were similar for gestational age (GA), birth weight (BW), prolonged rupture of membranes (PROM), Apgar scores, IVH, and aNRBC counts. PVL correlated significantly with low partial pressure of CO2 (pCO2) and IVH (P < 0.01). In logistic regression, when GA, gender, PROM, antenatal steroid therapy, 1 (or 5) minute Apgar scores, IVH grade, nosocomial sepsis, patent ductus arteriosus, necrotizing enterocolitis (NEC), need for pressors, aNRBC counts and lowest pCO2 were used as independent variables, pCO2 (P = 0.002), IVH grade (P = 0.001), GA (P = 0.038), NEC (P = 0.061) and use of dopamine (P = 0.010) remained in the analysis (total R2 = 68.2%).
Conclusions: In contrast to severe IVH, aNRBC counts do not predict the development of PVL.