Appropriateness of prescriptions of recommended treatments in OECD - - PowerPoint PPT Presentation

Appropriateness of prescriptions of recommended treatments in OECD health systems: findings based on the Long-Term Registry of the ESC

• n Heart Failure

Aldo P. Maggioni, Kees Van Gool, Nelly Biondi, Renato Urso, Niek Klazinga, Roberto Ferrari, Nikolaos Maniadakis and Luigi Tavazzi

Disclosures: none

Background

• HF affects approximately 2% to 3% of the population,

rising to around 10% in patients aged =>70 years

• HF is a common reason for hospital admissions: 14% of all

CV hospital admissions in Organisation for Economic Co-

• peration and Development (OECD) countries
• Survival of patients with chronic HF improved over the last

2 decades due to the widespread adoption of treatment recommended by current clinical practice guidelines

• However, guidelines are often adopted too slowly or are

applied partially and inconsistently

• The structure and organization of the health systems is

likely to play an important role in explaining the insufficient application of guideline recommendations in drug prescriptions

Aim of the study

• To identify which patient clinical characteristics

and which health system characteristics are associated with incomplete guideline incorporation and application in clinical practice

Methods (1)

• Patient characteristics are sourced from the ESC Heart

Failure Long-Term Registry

• Country-level data are derived from the OECD’s Health

System Characteristics Survey, the OECD Health Statistics 2013 Database and Eurostats’ Statistics on Income and Living Conditions (EU-SILC)

• For the purpose of this analysis, only OECD countries, for

which health system characteristics data is available, were considered

• Patients hospitalized for acute HF and chronic HF patients

with an EF =>40% were excluded since current guidelines do not include EB recommendations for these patients

• Local audits were planned to check in a randomized sample
• f centres and patients the quality of the collected data and

the consecutiveness of enrolment

Methods (2)

• Inappropriate prescription of pharmacological

treatments recommended by ESC guidelines was defined as follows:

q Patients not treated at all with at least one of the two

recommended treatments (ACE-Inhibitors/ARBs and beta- blockers)

• r

q Patients treated with both ACE-Inhibitors/ARBs and beta-

blockers, but with a suboptimal dosage and

q Absence of a documented contraindication or intolerance to

the two recommended classes of drugs

Methods (3)

• Statistical analysis

q First step:

§ Each patient-level and country-level variables, either continuous or categorical was examined for its univariate association with inappropriate prescriptions

q Second step:

§ Given the hierarchical nature of the data (patients nested within countries), a hierarchical model was tested in an empty model with the country identifier as random intercept § All patient-level variables were then included in the multivariable model § Thereafter, country group level covariates were added and examined one at a time

q The association between inappropriate prescription and each covariate

was calculated using odds ratios (OR) with 95% confidence intervals (CI)

§ The data were analysed using the SAS, version 9.3 for Windows statistical software (SAS institute, Cary, NC, USA)

Patient disposition

21 Countries 211 Cardiology centers From May 2011 to Dec 2013

• n. 17,901 pts

Total number

• n. 572 pts

Sweden

• n. 3,357 pts

Hospitalised HF pts

• n. 4,605 pts

HF reduced EF

• n. 2,352 pts

HF with EF ≥40%

• n. 1,303 pts

Missing EF data

• n. 12,189 pts

OECD countries

OECD: Organisation for Economic Co-operation and Development; HF: heart failure; EF: ejection fraction

• n. 11,617 pts

Pts with available adherence data

• n. 8,260 pts

Chronic HF pts

• n. 5,712 pts

non-OECD countries

Inappropriate prescriptions by OECD countries

• N. of

centres

• N. of Pts

Inappropriate drug prescription (%) Austria 4 126 17.5 Czech Republic 7 337 21.4 Denmark 21 160 38.1 Estonia 2 21 76.2 France 12 322 15.8 Greece 4 50 76.0 Hungary 5 239 15.9 Israel 2 286 18.2 Italy 17 754 28.8 Poland 17 236 32.2 Portugal 10 354 9.6 Slovak Republic 4 117 15.4 Slovenia 8 67 22.4 Spain 20 1440 25.3 Turkey 5 96 24.0 Total 138 4605

23.8

Inappropriate prescriptions of participating centres, within and across country variation

.2 .4 .6 .8 1 Proportion of patients not adherent A U T C Z E D N K E S P E S T F R A G R C H U N I S R I T A P O L P R T S V K S V N T U R

Panel A: All centres

.2 .4 .6 .8 1 Proportion of patients not adherent A U T C Z E D N K E S P E S T F R A G R C H U N I S R I T A P O L P R T S V K S V N T U R

Panel B: Centres with sample of 10 or more patients

Patients’ characteristics and ORs for inappropriate prescriptions estimated by univariate models

Appropriate (n=3508) N (%) Inappropriate (n=1097) N (%) P-value Age (years) <65 65-75 >75 1857 (52.9) 1008 (28.7) 643 (18.3) 523 (47.7) 318 (29.0) 256 (23.3) 0.001 Men 2752 (78.5) 844 (76.9) 0.31 Body mass index (kg/m2) <22 22-25 >25 327 (9.3) 695 (19.9) 2475 (70.8) 91 (8.4) 228 (20.9) 770 (70.7) 0.50 Ejection fraction <30% 2001 (57.0) 570 (52.0) 0.04 NYHA class III or IV 962 (27.4) 269 (24.5) 0.06 Heart rate >70 bpm 1577 (45.0) 486 (44.3) 0.73 Systolic Blood Pressure (mmHg) <110 110-130 >130 1227 (35.0) 1399 (39.9) 881 (25.1) 365 (33.3) 435 (39.6) 297 (27.1) 0.37

Patients’ characteristics and ORs for inappropriate prescriptions estimated by univariate models

Appropriate (n=3508) N (%) Inappropriate (n=1097) N (%) P-value Creatinine >1.5 mg/dL 644 (19.6) 216 (21.2) 0.28 Sodium <136 mEq/L 367 (11.6) 121 (12.5) 0.48 Ischemic etiology 1664 (47.4) 558 (50.9) 0.05 Atrial fibrillation 1186 (33.8) 378 (34.5) 0.72 Left bundle branch block 733 (23.0) 234 (24.9) 0.23 Mitral regurgitation 1032 (29.9) 368 (34.9) 0.003 COPD 560 (16.0) 145 (12.9) 0.01 Peripheral arterial diseases 412 (11.8) 144 (13.2) 0.24 Chronic kidney diseases 638 (18.3) 221 (20.4) 0.14 Diabetes mellitus 1174 (33.5) 378 (34.5) 0.57 Prior stroke 322 (9.2) 107 (9.8) 0.60 Depression 266 (7.6) 69 (6.3) 0.17 COPD=chronic obstructive pulmonary disease

Country level variables and ORs for inappropriate prescriptions estimated by univariate models

Appropriate (n=3508) Inappropriate (n=1097) P- value ACCESS

Percentage of population skipping a doctor consultation 6.3±3.2 6.9±3.2 <.0001 Patient sample living in countries where health services are primarily free at the point of care 1874 (55.0) 613 (56.8) 0.29 Out of pocket expenditures on medical goods (per capita US$) 213±37 210±31 0.04 Out of pocket expenditures on

• utpatient care (per capita US$)

215±100 230±104 <.0001

• N. of annual doctor consultations (per

capita) 7.4±2.2 7.3±2.0 0.02

Country level variables and ORs for inappropriate prescriptions estimated by univariate models

Appropriate (n=3508) Inappropriate (n=1097) P- value RESOURCES

Percentage of total health expenditure spend on ambulatory care 24.5±7.8 23.0±6.8 <.0001 Number of GPs (per 1000 population) 0.9±0.5 0.8±0.4 <.0001 Patients living in countries where GPs are mostly privately employed 975 (33.4) 261 (27.2) <0.001 Patients living in countries where GPs are primarily paid fee-for-service 305 (11.7) 60 (6.5) <.0001 Patient sample living in countries where there are obligation or incentives to register with a GP 2981 (89.5) 928 (89.2) 0.81

Country level variables and ORs for inappropriate prescriptions estimated by univariate models

Appropriate (n=3508) Inappropriate (n=1097) P- value QUALITY

Patients living in countries with formal accreditation requirements for primary care practices to operate 2357 (70.8) 688 (66.2) 0.01 Patient sample living in countries with formal system of continuous medical education 2900 (87.1) 907 (87.2) 0.90 Patients living in countries with use of a patient registration system 3171 (95.2) 964 (92.7) <0.001 Patients living in countries with electronic exchange of information between providers 2381 (71.5) 694 (66.7) <0.001 Patients living in countries with incentives to comply with treatment guidelines 2851 (88.4) 847 (83.2) <.0001

Odds ratios by the multi-level logistic model

OR 95%CI P-value Age (years) <65 65-75 >75 Ref 1.0 1.2 0.8-1.3 0.9-1.5 0.87 0.19 Men 0.9 0.7-1.2 0.57 Body mass index (kg/m2) <22 22-25 >25 0.9 Ref 0.9 0.6-1.2 0.7-1.1 0.37 0.21 Ejection fraction <30% 0.9 0.7-1.1 0.13 NYHA class III or IV 0.9 0.7-1.1 0.15 Heart rate >70 bpm 1.1 0.9-1.3 0.41 Systolic Blood Pressure (mmHg) <110 110-130 >130 0.9 Ref 1.0 0.7-1.1 0.8-1.3 0.20 0.99 Creatinine >1.5 mg/dL 1.1 0.8-1.5 0.43 Sodium <136 mEq/L 1.2 0.9-1.6 0.26

Odds ratios by the multi-level logistic model

OR 95%CI P-value Ischemic etiology 1.2 1.0-1.4 0.12 Atrial fibrillation 1.1 0.9-1.4 0.22 Left bundle branch block 1.2 1.0-1.6 0.07 Mitral regurgitation 1.4 1.1-1.7 0.01 COPD 0.7 0.5-0.9 0.01 Peripheral arterial diseases 1.1 0.8-1.5 0.44 Chronic kidney diseases 0.9 0.7-1.2 0.51 Diabetes mellitus 1.1 0.9-1.3 0.51 Prior stroke 1.2 0.9-1.7 0.20 Depression 0.9 0.6-1.2 0.36 Incentives to comply with treatment guidelines 0.4 0.2-1.1 0.07 COPD=chronic obstructive pulmonary disease

Limitations

• The analysis is limited by the small number of patients in some

countries which restricts our ability to explore country-level data

• Consecutive enrolment cannot be fully proven. In the attempt to

get the registry burden feasible and realistic, the enrolment was done

• n a “one day per week” basis. Local audits were performed to check

quality of data and consecutiveness

• To maximize representativeness, the centres were selected in

proportion to the size of the population of the countries and taking into account the technological levels of the cardiology centres

• The patients were enrolled by cardiologists and therefore the study

population may not be representative the entire chronic HF population

• In a relevant rate of patients (16%) EF was not measured and

these patients were excluded from the analysis. This can be considered an inappropriate way to manage HF patients, that probably merits future specific evaluations

Conclusions

• ESC-OECD analysis shows:

q national borders can explain some variations in treatment

patterns

q the importance of within country variation

• Lack of appropriateness of prescriptions could impact

patient outcomes and health care costs

• More efforts should be made towards understanding the

role of health system characteristics in explaining application of guidelines both within and between countries

• Such an understanding could aid decision makers to

design more effective policies to improve quality of health care and to potentially improve patients’ outcomes and reduce costs

EURObservational Research Programme

19

EORP Sponsor Meeting –Brussels – 16 April 2014

19

Back up

EURObservational Research Programme

20

EORP Sponsor Meeting –Brussels – 16 April 2014

20

Methods (2)

• Inappropriate prescription of pharmacological treatments

recommended by ESC guidelines was defined as follows:

q Patients not treated at all with at least one of the two recommended

treatments (ACE-Inhibitors/ARBs and beta-blockers)

• r

q Patients treated with both ACE-Inhibitors/ARBs and beta-blockers, but

with a suboptimal dosage and

q Absence of a documented contraindication or intolerance to the two

recommended classes of drugs

• Clinical variables

q Demographic characteristics (age and gender) q Vital signs and symptoms (SBP, HR, NYHA class) q Etiology of HF q Laboratory and instrumental results (LBBB at ECG, EF, sodium,

creatinine)

q Classical risk factors (BMI, diabetes mellitus, hypertension) q Comorbid condition and vascular history (history of stroke, AF, mitral

regurgitation, depression, COPD, chronic kidney dysfunction, PAD)

Methods (3)

• OECD Health System Characteristics Survey

q Patient access to health care q Financial and physical resources available for health care q Quality improvement initiatives within a health care system

• OECD Health and Eurostat Statistics

q Health resources and activities q Health expenditures

• Eurostat’s Statistics on Income and Living Conditions

(EU-SILC)

q Percentage of people who reported that they did not visit their

doctor in the previous 12 months when they felt they needed health care

Methods (2)

• Inappropriate prescription of pharmacological treatments

recommended by ESC guidelines was defined as follows:

q Patients not treated at all with at least one of the two recommended

treatments (ACE-Inhibitors/ARBs and beta-blockers)

• r

q Patients treated with both ACE-Inhibitors/ARBs and beta-blockers, but

with a suboptimal dosage and

q Absence of a documented contraindication or intolerance to the two

recommended classes of drugs

• Clinical variables

q Demographic characteristics (age and gender) q Vital signs and symptoms (SBP, HR, NYHA class) q Etiology of HF q Laboratory and instrumental results (LBBB at ECG, EF, sodium,

creatinine)

q Classical risk factors (BMI, diabetes mellitus, hypertension) q Comorbid condition and vascular history (history of stroke, AF, mitral

regurgitation, depression, COPD, chronic kidney dysfunction, PAD)