Background: Health anxiety (Hypochondriasis) is prevalent, but rarely diagnosed and treated. The essential features of health anxiety are exaggerated rumination with intrusive worries about harbouring a serious illness. Severe health anxiety might be persistent and associated with severe psychological and physiological impairment. Treatment of health anxiety is sparsely investigated. Acceptance and Commitment Therapy (ACT) is a new third-wave behavioural cognitive therapy that has shown a positive effect in the treatment of mood and anxiety disorders.
Objective: To examine the effect of ACT in groups for patients with severe health anxiety.
Methods: 126 patients consecutively referred from primary care physicians or hospital departments, meeting research criteria for severe health anxiety, were block-randomised to either; a) ACT in groups or b) a ten- month waiting list. Primary outcome was self-rated improvement in illness worry on the Whiteley-7 Index (WI) 10 months after randomisation. Secondary outcomes were improvement in emotional distress (SCL-8) and physical symptoms (SCL-somatisation subscale) at ten months after randomisation.
Preliminary results: The intention-to-treat analysis showed that patients in the ACT group improved 22.1 score points on the WI at the primary end point, which were significantly better than the waiting list control group (unadjusted mean difference 21.3, 95% CI 12.6 to 30, p<0.001) and effect sizes were large (d=0.89, 95% CI 0.50 to1.29). The ACT group also
improved significantly more than the control group on most secondary outcomes.
Conclusion: ACT group therapy seems feasible, acceptable and effective in treatment of severe health anxiety.
O05.03 Eva Bjerre Ostenfeld
PREADMISSION USE OF GLUCOCORTICOIDS AND ANASTOMOTIC LEAKAGE FOLLOWING COLORECTAL CANCER RESECTION: A DANISH POPULATION-BASED COHORT STUDY
E.B. Ostenfeld1, 2, R. Erichsen1, J. Baron3, O. Thorlacius-Ussing2, L.H. Iversen4, A.H. Riis1, H.T. Sørensen1
1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark, 2Department of Surgery (A), Aalborg University Hospital, Aalborg, Denmark, 3Department of Medicine, University of North Carolina School of Medicine, Chapel Hill, NC, USA., 4Department of Surgery (P), Aarhus University Hospital, Aarhus, Denmark
Background: It is controversial whether glucocorticoid use affects of anastomotic failure.
Methods: We conducted a population-based cohort study using Danish medical registries to examine the association between preadmission glucocorticoid use and anastomotic leakage following colorectal cancer (CRC) surgery. All patients treated with a primary anastomosis (2001-2011) following a CRC resection were included. Subjects who filled their most recent glucocorticoid prescription ≤90, 91-365, and >365 days before their surgery date were characterized as current, recent, and former users, respectively. We computed absolute risk of anastomotic leakage within 30
61
days after CRC resection and odds ratios (ORs), using logistic regression models and adjusting for potential confounders.
Results: Of the 24 043 CRC patients included, 2679 (11.1%) filled at least one prescription of glucocorticoids in the year preceding their surgery date.
Overall, 1779 (7.4%) anastomotic leakages occurred. Absolute risk of leakage among current users of oral glucocorticoids was 9.0% versus 7.3%
among non-users (OR = 1.50, 95% CI: 1.05-2.13). Risk estimates for recent and former use of oral glucocorticoids were slightly lower, yet statistically imprecise, as were estimates for current use of inhaled glucocorticoids and use of intestinal-acting glucocorticoids.
Conclusion: Current use of glucocorticoids was associated with an increased relative risk of anastomotic leakage following CRC resection.
However, the increase in absolute risk was small.
O05.04 Dorthe Krogsgaard Bonnerup
PRESCRIBING ERRORS IN ACUTELY ADMITTED MEDICAL PATIENTS D.K. Bonnerup1, 3, M. Lisby2, E.A. Sædder3, A. Eskildsen1, L.P. Nielsen3
1Department of Biomedicine – Centre for Clinical Pharmacology , Aarhus University, 2Research Centre of Emergency Medicine, Department of Clinical Medicine, Aarhus University, 3Department of Clinical
Pharmacology, Aarhus University Hospital
Background: An algorithm that can stratify patients according to risk of medication errors have been developed. In order to test the algorithm in a randomised controlled trial, the number of prescribing errors is vital to know.
We define prescribing errors as errors in prescribing causing harm or implying a risk of harming the patients.
Aim: The aim of the study was to assess the number of prescribing errors in acutely admitted adult patients at the Emergency Department at Aarhus University Hospital, Denmark.
Methods: Patients were included prospectively on admission to the Emer-gency Department at Aarhus University Hospital. Patients aged 18 years or older who received at least one drug prior to admission were eligible for inclusion. Suicidal and intoxicated patients were excluded. A pharmacist and a clinical pharmacologist assessed medical records after discharge for prescribing errors during the hospitalization. These assessments were finally validated independently by two clinical pharmacologists. Chi square tests were used to assess factors related to risk of prescribing errors.
Results: 103 patients were included of which 51 patients (49.5%) experienced one or more prescribing errors (1-6). 86 prescribing errors were found in 748 prescriptions, corresponding to 9.5 percent of
prescriptions. Drug-drug interactions (N=19) and dose-related errors (N=18) were the most frequent prescribing errors seen. Age above 65 years (p=0.013) and reduced kidney function (p=0.03) were factors related to increased risk of prescribing errors.
Conclusion: Prescribing errors were found in 9.5 % of prescriptions (86 errors in 748 prescriptions).
62
O05.05 Merete Gregersen
A LIBERAL BLOOD TRANSFUSION STRATEGY IMPROVES SURVIVAL IN NURSING HOME RESIDENTS WITH HIP FRACTURE
M. Gregersen1, L.C. Borris2, E.M. Damsgaard1
1Department of Geriatrics, Aarhus University Hospital, 2Department of Orthopaedic Surgergy, Aarhus University Hospital
Introduction: Surgery blood loss leading to anemia is common in elderly hip fracture (HF) patients. Nursing home residents are frail elderly with high mortality risk after HF. The most optimal transfusion threshold in the frail elderly HF patients has not yet been examined. Our aim was to test the hypothesis that a more liberal transfusion strategy would improve survival in the frail anemic elderly who had undergone surgery for HF.
Material and methods: In a randomized controlled trial, we included 65+
years old nursing home residents hospitalised with HF at the Department of Orthopaedic Surgery, Aarhus University Hospital. After surgery, we
randomly assigned the residents to either a liberal transfusion strategy (a hemoglobin threshold of 7 mmol/L [11.3 g/dL]) or a restrictive blood transfusion strategy (a hemoglobin threshold of 6 mmol/L [9.7 g/dL]).
Within the first week after surgery, hemoglobin was measured 4-6 times, and at least once a week during the following three weeks. When a blood test indicated blood transfusion, it was administered within 24 hours, and only one blood unit before a new measurement. Outcome was time to death within 90 days and was analysed in a Cox proportional hazard model.
Results: Hundred and sixty nursing home residents were enrolled. The intention-to-treat analysis showed that the 90-day mortality rate was 20%
in the residents treated by the liberal strategy versus 35% in the residents treated by the restrictive strategy (p-value=0.03). Hazard ratio was 0.53 (95% CI: 0.29; 0.98).
Conclusion: Survival is improved by complying with a liberal transfusion strategy in the anaemic nursing home residents with hip fracture.
O05.06 Mikkel Andreas Strømgaard Andersen
FIRST HOUR QUINTET 1-1-2 CALLERS AND COMORBIDITY
M.S. Andersen1,2, S.P. Johnsen2, J.N. Sørensen3, S.B. Jepsen4, J.B. Hansen1, E.F.
Christensen1
1Research Department, Prehospital Emergency Medical Services, Central Denmark Region,2Department of Clinical Epidemiology, Aarhus University Hospital ,3Emergency Medical Communication Center, Capital Region of Denmark,4Emergency Medical Communication Center, Odense University Hospital, Region of Southern Denmark
Introduction: First Hour Quintet (FHQ) consists of cardiac arrest, angina pectoris, stroke, severe breathing difficulties and severe trauma - all patients for whom the first hour after debut of symptoms is important. The task for the new nurse, paramedic and doctor staffed emergency medical communication centers (EMCC) in Denmark is to dispatch high priority ambulances to high-risk patients. We aimed to study the impact of
co-63
morbidity on the risk of ICU treatment and death among FHQ 1-1-2 callers.
Methods: We did a register-based follow-up study on 1-1-2 callers in a six- month period. The patients were included if they belonged to the FHQ groups in the dispatch protocol. The Charlson Comorbidity Index was used to categorize comorbidity and constructed for each patient using hospital discharge diagnoses. Logistic regression was used to analyze the
association between comorbidity and risk of ICU treatment or death on same or the following day after a 1-1-2 call.
Results: A total of 20,500 patients were included. Of these, 11,006 patients had comorbidity. In total 1,818 patients received ICU treatment or died.
Odds ratio (OR) for ICU treatment or death for patients with mild liver disease was 1.59 (95 % CI: 1.18-2.13 P< 0.01), distal vascular disease 1.33 (95 % CI: 1.13-1.56 P< 0.01), heart failure 1.32 (95 % CI: 1.14-1.53 P< 0.01), Chronic Obstructive Pulmonary Disease 1.29 (95 % CI: 1.14-1.46. P< 0.01), any tumor 1.18 (95 % CI: 1.01-1.38. P= 0.04). No other comorbidities were associated with risk of ICU treatment or death for FHQ patients.
Conclusion: Comorbidity seems to have a significant impact on the risk of an adverse outcome for first-hour quintet 1-1-2 callers.
64