A374
VA L U E I N H E A LT H 1 8 ( 2 0 1 5 ) A 3 3 5 – A 7 6 6
PCV2
ADVERSE DRUG REACTIONS, MEDICATION ADHERENCE AND PHYSICIANS’
PRESCRIBING BEHAVIOR: WHICH RELATIONSHIP? A DATABASE ANALYSIS ON
STATIN USE
Leporini C1, Degli Esposti L2, Sangiorgi D2, Ursini F1, Scicchitano F1, Russo E1
1University of Catanzaro, Catanzaro, Italy, 2CliCon S.r.l., Ravenna, Italy
OBJECTIVES: Pharmacological adherence plays an essential role in controlling hypercholesterolemia. Pharmacological care is also affected by the prescribing inappropriateness and adverse drug reactions (ADRs). ADRs can lead to a perceived lack of
therapy effectiveness and subsequent suboptimal adherence. Our aim was to evaluate
the impact of statin-induced, muscle-related ADRs on patients’ therapy adherence,
and how this relationship may compromise the efficacy of physicians’ prescribing
choices and health outcomes in clinical practice METHODS: A retrospective cohort
study was performed using data from databases of 4 Local Health Units (LHUs) located
in Emilia-Romagna, Toscana and Umbria regions, with an overall population of about
1.1 millions of inhabitants. All subjects aged≥18 years with a first prescription for
statins in the period 01/01/2007-30/06/2008 were included. Baseline and follow-up
LDL-C levels were considered RESULTS: A total of 71,855 patients (51% men, age
68.6±10.6) were included. Among them, 31,544 (43.9%) had ≥ 1 LDL-C values during
the follow-up period: only 37.4% of these patients achieved LDL-C target. 23.6% of
patients underwent ≥1 CPK measurement: out of range values were identified in
37.8% of them. Patients with steadily normal values of CPK showed a better medication-taking behavior compared to patients whose CPK values were out of normal
range. Furthermore, a greater percentage of switching to other statin drug and/or
dosage was observed in patients with out of range CPK levels compared to those with
normal serum CPK levels CONCLUSIONS: Our findings indicate that physiological
serum CPK levels were associated with greater adherence to statin therapy, supporting a causal link between non-adherence behavior and muscle-related ADRs
occurrence. Since statin benefits are associated with their chronic use, physicians
should be aware about the relevance of monitoring their patients for this harmful
link to prevent unsuitable therapeutic decisions, further decreasing adherence, and
achieve long-term health outcomes
PCV3
LEAD DAMAGE IN THE FIRST-YEAR FOLLOWING CARDIAC IMPLANTABLE
ELECTRONIC DEVICE (CIED) REPLACEMENT: INCIDENCE AND COSTS BASED ON
DEVICE TYPE
Nichols CI1, Vose JG1, Mittal S2
1Medtronic Advanced Energy, Portsmouth, NH, USA, 2Valley Health System of New York and New
Jersey, New York, NY, USA
OBJECTIVES: This study evaluated the incidence and costs of lead damage within one
year of pacemaker (PM), implantable cardiac defibrillator (ICD), and cardiac resynchronization therapy defibrillator (CRT-D) device replacement. METHODS: Using Truven
US insurance claims (1/1/2009 – 12/31/2013), we identified patients who underwent
a CIED generator replacement. We excluded patients with existing lead damage or
active infection. Data were analyzed for one year prior to and following the replacement procedure, with device replacement the index date for analysis. During followup we considered lead damage to have occurred if there was any recorded visit with a
procedure code for lead revision or repair and no concurrent code for infection. A Cox
model, adjusted for baseline demographics and comorbidities, was used to evaluate
variables associated with risk of lead damage. RESULTS: The study cohort included
45,250 patients (72±14 years; 64% male), including 22,557 (50%) PMs, 20,632 (46%)
ICDs, and 2,061 (5%) CRT-Ds. Lead damage was observed in 406 (0.90%) patients. The
incidence of lead damage was significantly influenced by CIED type (PM: 0.46%; ICD:
1.27%; CRT: 1.94% p < .001). Similarly, total hospitalization cost associated with lead
revision / repair was influenced by CIED type (PM: $19,959; ICD: $ 24,885; CRT: $46,229,
p=0.048). As compared to PMs, ICDs (HR = 2.14, 95% CI 1.68 – 2.74) and CRTs (HR = 2.97,
95% CI 1.99 – 4.43) had a significantly higher risk of lead damage. CONCLUSIONS:
This is the first nationally representative study to assess the incidence and associated costs of lead damage following CIED generator replacement in the US. Patients
with an ICD or CRT have a significant 1-2% risk of lead damage, which is associated
with substantial hospitalization-related costs. Strategies are needed at the time of
generator replacement to minimize inadvertent lead damage.
PCV4
NETWORK META-ANALYSIS OF ORAL ANTICOAGULANTS FOR PRIMARY
PREVENTION, TREATMENT AND SECONDARY PREVENTION OF VENOUS
THROMBOEMBOLIC DISEASE, AND FOR PREVENTION OF STROKE IN ATRIAL
FIBRILLATION
Lopez-Lopez JA1, Sterne J1, Bodalia PN2, Bryden PA1, Davies P1, Okoli GN1, Thom H1,
Caldwell DM1, Dias S1, Eaton D3, Higgins J1, Hollingworth W4, Salisbury C1, Savovic J1,
Sofat R5, Stephens-Boal A6, Welton NJ1, Hingorani A5
1University of Bristol, Bristol, UK, 2University College London Hospitals, London, UK,
3Anticoagulation Europe, Kent, UK, 4University of Bristol, BRISTOL, UK, 5University College
London, London, UK, 6Thrombosis UK, LLANWRDA, UK
OBJECTIVES: To determine the best oral anticoagulant(s) across four medical conditions: stroke prevention in atrial fibrillation (AF), primary prevention of venous
thromboembolic disease (VTE), acute treatment of VTE, and secondary prevention of
VTE; using network meta-analysis to compare novel oral anticoagulants (NOACs) with
warfarin and/or low molecular weight heparin (LMWH). METHODS: We undertook
four systematic reviews and network meta-analyses of randomised controlled trials.
An extensive search and risk of bias assessment were followed by network meta-analyses within each condition, and additionally we performed combined meta-analyses
across the four reviews for key safety outcomes. Through consultation between clinicians and methodologists, we needed to make various decisions about groupings of
interventions and doses, and of different endpoints, for analysis. RESULTS: 85 trials
were included across the reviews, including a mixture of phase III trials and phase
II dose-ranging trials. For stroke prevention in AF, apixaban (5mg bd) was among the
best interventions for a range of efficacy and safety outcomes with mean rank 2.35
in relation to stroke or systemic embolism and mean rank 1.15 for major bleeding.
Edoxaban (60mg od) was also highly ranked (means 3.24 and 1.92, respectively). We
did not see strong evidence that NOACs have advantages over LMWH in primary
prevention, or over warfarin in acute treatment and secondary prevention of VTE, but
risk of bleeding complications was lower for some NOACs than for warfarin. In sensitivity analyses, key decisions we made in the process did not materially affect the
results. CONCLUSIONS: NOACs have advantages over warfarin for stroke prevention
in AF and acute treatment of VTE, but we found no evidence that they should replace
the standard treatments for primary and secondary prevention of VTE.
PCV5
ASSESSMENT OF A PREDICTIVE RISK MODEL FOR CLASSIFYING PATIENTS WITH
MULTIMORBIDITY IN THE BASQUE COUNTRY
Millán E1, Lopez Arbeloa G1, Samper R1, Carneiro M1, Pocheville E1, Aurrekoetxea J2, de
Pablos I2, Orueta J1
1Osakidetza, Basque Health Service, Vitoria-Gasteiz, Spain, 2Basque Health Department, VitoriaGasteiz, Spain
OBJECTIVES: To describe the rate of unplanned admissions in patients with multimorbidity and calculate the percent of patients with at least one unplanned admission by risk classification based on the results of a predictive model METHODS:
Cross-sectional study. In September 2013 all Basque Country citizens were classified
in risk groups based on a prediction model that used data from a 2 year period:
age, sex, diagnoses, procedures, prescriptions and costs from the first year as the
explanatory variables and healthcare cost in the second year as the dependent
one. A Patient with multimorbidity was defined as following: 2 or 3 from diabetes,
chronic obstruction pulmonary disease and chronic heart failure; or multimorbidity definition used by Ollero et al published elsewhere. People younger than 14 or
with a diagnosis of cancer, a transplant or dialysis were excluded. Multimorbidity
was classified in three risk groups based on the expected healthcare cost percentiles (< 75; 75-95; > 95). Case-mix Adjusted Clinical Groups (ACG) was used for both
prediction and definition of multimorbidity. Rate of unplanned admissions and the
probability of one or more unplanned admissions by risk group during the following year to their classification were calculated RESULTS: A total of 24,247 people
met the criteria of multimorbidity (61.3%males) with a mean age of 75.5 years.
2,795 were classified as low-risk, 10,598 as medium-risk and 10,853 as high-risk.
The rate of unplanned admissions was 148.4/1000 in the low-risk group, 340.3/1000
in the medium-risk group and 861.6/1000 in the high-risk group (Mantel-Haenszel
trend test p< 0,001). 11.4%, 22.8% and 44.6% of the patients in each risk group
required one or more unplanned admissions during the following year, respectively
(p< 0,001) CONCLUSIONS: Predictive risk models, based on ACG, allow classifying
patients with multimorbidity in risk groups. This makes possible a better development of personalized, cost-efficient and patient-centered care plans
PCV6
CLINICAL AND ECONOMIC IMPACT OF CHRONIC HEART FAILURE IN GENERAL
PRACTICE: ANALYSIS OF 13.633 PATIENTS
Lapi F1, Cricelli I1, Simonetti M1, Colombo D2, Nica M2
1Health Search, Italian College of General Practitioners and Primary Care., Firenze, Italy, 2Novartis
Farma, Origgio, Italy
OBJECTIVES: Update epidemiology and estimate Chronic Heart Failure (CHF) related
costs in general practice setting METHODS: We used Health Search, an Italian general practice database formed by about 1,000 GPs. We collected demographics, clinical informations, drug prescriptions, co-morbidities, and lifestyle measurements,
reflecting “real world” practice. RESULTS: We analyzed data of 13,633 CHF (mean
age was 76.78 years) patients that resulted in a prevalence of 1.25%. We observed
similar prevalence in men and women; prevalence increased significantly with age,
in both men and women. There were some substantial differences across regions
(eg.: Emilia Romagna 1.54% vs. Sardinia 0.61%). Incidence rate was slightely lower
than that reported in the literature (1.99*1000 person-years) with some regional
diffrences: 3.24*1000 Emilia Romagna as oposed to 1.21*1000 in Puglia and Marche.
Women had a higher incidence rate than men (2.05*1.000 vs. 1.92*1000). Overall,
mean cost in a cohort of 7.710 patients (excluding 67,7% of patients that had a null
cost registrated by phisicians) was 950.22€ /patient/year (DS: 1135,27), while median
cost was 643,16€. Age and NYHA class seemed to influence costs for elderly patients,
while ex-smokers, alcohol abusers and patients with alcohol related diseases had a
significant reduction of costs. CONCLUSIONS: Real world data is needed to understand the epidemiology of and costs due to CHF in primary care. The epidemiology
of CHF here reported is in line with current literature. Concerning CHF-related costs
and the respective determinants, this is the first study examining the Italian general
practice setting. The present analysis gives an important contribution in updating
epidemiology and estimating CHF-related costs.
PCV7
SYSTEMATIC LITERATURE REVIEW OF THE BURDEN OF ILLNESS IN
HYPERTRIGLYCERIDEMIA
Martin AL1, Travers KU1, Burns MD1, Palmer M2, Henriksson KM3, Rikner K3
1Evidera, Lexington, MA, USA, 2AstraZeneca, Melbourn, UK, 3AstraZeneca, Mölndal, Sweden
OBJECTIVES: Hypertriglyceridemia (HTG) is a common condition characterized
by elevated triglyceride (TG) levels with the threshold inconsistently defined but
often appearing as ≥ 2.26 mmol/L; patients with TG levels ≥ 5.6 mmol/L have severe
HTG (sHTG). The objective of this study was to evaluate the burden of HTG and
report unmet needs. METHODS: MEDLINE- and Embase-indexed databases were
systematically searched, pairing keywords for elevated TG levels with outcomes, to
identify English-language articles published from 1/1/2002–4/14/2014. Conference
proceedings from the past four years were also reviewed. Studies evaluating clinical, patient-centered, economic, and epidemiologic outcomes in adults with TG
levels ≥ 2.26 mmol/L were included. RESULTS: Of the 1,225 articles identified, 160
studies were included. Prevalence of HTG (1.2%–55.8% globally) was reported inconsistently. Significant associations were identified between HTG and cardiovascular