58

European Journal of Public Health, Vol. 24, Supplement 2, 2014

Downloaded from http://eurpub.oxfordjournals.org/ by guest on January 16, 2017

D.2. Health care and inequalities in chronic diseases The impact of reimbursement systems on equity in access, utilization and quality of primary health care: a systematic review Wenjing Tao W Tao, J Agerholm, M Weinryb, B Burstro¨m Centre for Epidemiology and Community Medicine, Stockholm County Council, Stockholm, Sweden Contact: [email protected]

Background Different reimbursement systems might result in differences in access, utilization and quality of health care in relation to ethnicity and socioeconomic status. Previous reviews assessing the impact of reimbursement systems on health care have not taken equity aspects into account. This study aims to provide a comprehensive overview of the impact of reimbursement system on socioeconomic and ethnic patterns in access, utilization and quality of primary health care.

Methods A systematic search was conducted in MEDLINE and Web of Science for English and Swedish literature published between January 1980 and September 2013, supplemented by reference tracking. Two authors independently appraised the selected articles and extracted relevant data. Results 24 studies were identified from the Web of Science, of which 19 were from the United Kingdom and assessed the impact of pay-for-performance on utilization and quality of health care. Five studies on other reimbursement systems were conducted in the United States, Canada and Norway. The main sources of data were surveys and administrative databases. Socioeconomic status and ethnicity were the most common aspects examined. Pay-for-performance seems to have some effect on reducing differences in utilization and quality of health care between socioeconomic-, ethnic- and

7th European Public Health Conference: Thursday 20 November 2014 16:00-17:00

age-groups, but the gaps largely persisted after introduction of this system. The effects were most apparent for chronic diseases, but the results varied between studies. The studies comparing capitation and fee-for-service were few and varied greatly in measured outcomes. Capitation seemed to increase health care access overall compared to fee for service, but the increase was most apparent for patients from the higher income quartiles and did not reduce the gap between socioeconomic groups. However, differences in health care access between ethnic groups seemed to be reduced by capitation. Conclusions The empirical studies assessing the impact of reimbursement system on inequity in health care were few and heterogeneous, and suggest that the type of reimbursement system might have some impact on equity in health care, but the effects were small and not convincing. Key messages  Type of reimbursement system might have some impact on equity in health care, but the effects are small.  Studies assessing the impact of reimbursement system on inequity in health care are few and heterogeneous.

Ileana Manoela Prejbeanu IM Prejbeanu1, MG Mihai1, ML Cara2 1 Environmental Health Department, University of Medicine and Pharmacy of Craiova, Craiova, Romania 2 General Directorate of Social Assistance and Child Protection Dolj County, Craiova, Romania Contact: [email protected]

Background Unfortunately Romania comes first in the European Union as far as the mortality due to cervical cancer is concerning, its rate recording a continuous increase because of the advanced stages disease detection. In this context, we evaluated the women’s cognitive and attitudinal behavior in relation to the early detection of cervical cancer, through the application of a questionnaire on 617 women aged from 18 to 75, low-, middle- or high educated, living in one of the largest Romanian cities, respectively in surrounding rural. Results Statistic analysis of the answers indicates 43.3% of the subjects have not had a routine gynecological exam for at least seven years or even never; this situation was found in the subgroup of women from the rural environment, with an elementary educational level (p < 0.001). More than two thirds of the women have not had a Pap test for seven years or even never, residence area and education level being once again discrimination factors between the subgroups; the reasons usually explain this situation are negligence (23.8%), the lack of information on the existence of such an investigation (18.2%), the absence of a genital pathology (12.8%) or of a medical recommendation (11.8%). Conclusions Results suggest the need to provide uniform accessibility of the population to health services, by developing and implementing new strategies of educational and sanitary intervention mainly in rural communities. Key messages  Negligence and lack of information limit the women participation in having tests for cervical cancer early detection.  Educational interventions in rural communities are compulsory in order to decrease health inequalities in health and health care in this area. Regional and socioeconomic inequalities in lung cancer mortality in Belgium, 2001-2009 Paulien Hagedoorn P Hagedoorn, H Vandenheede, K Vanthomme, S Gadeyne Interface Demography, Department of Social Research, Faculty of Economic, Political and Social Sciences and Solvay Business School, Vrije Universiteit Brussels, Brussels, Belgium

Contact: [email protected]

Introduction Lung cancer is a leading cause of cancer death and lung cancer mortality among Belgian men is the highest in Europe. The association between lung cancer and socioeconomic status has been extensively studied, however, little research has focused on geographical differences in lung cancer mortality. Spatial analysis can provide insight into at risk regions and populations, which could aid health interventions, and can help to develop or prove hypothesis about key determinants of cancer mortality. Therefore, this study aims to examine regional and socioeconomic inequalities in Belgian lung cancer mortality in 2001-2009. Material and Methods Data from the Belgian National Mortality Database, a unique dataset linking census and register data to death certificates, are used. The study population comprises all inhabitants of the Belgian regions Flanders and Brussels during 2001-2009 aged 40+. Mortality by district is calculated using the age standardized mortality rate (ASMR), directly standardized to the 2001 population for Belgium. The role of individual socioeconomic status on lung cancer mortality is estimated using Poisson regression while controlling for age, marital status and health status. Results Lung cancer mortality for men and women aged 40+ is highest in the east of Flanders and is relatively low in the southeast. The exception is Brussels, where lung cancer mortality is relatively low for men but high for women. For men aged 40-64, higher lung cancer mortality is observed in the southwest of Flanders while lower mortality is found in the northeast. Overall, socioeconomic status has only minor effects on regional mortality differences. The effect of socioeconomic status on regional mortality is larger for women, and the largest effects are observed for the 40-64 age-category. Conclusion Both regional and socioeconomic inequalities in lung cancer mortality exist within Flanders and Brussels. Individual socioeconomic status can only partly explain the regional mortality differences. As indicated by previous studies, it is likely that other contextual factors are of influence as well. Future research should therefore look at additional determinants such as area-level socioeconomic status, access to healthcare and environmental factors as well. Key messages  Regional inequalities in lung cancer mortality persist after controlling for individual socioeconomic status.  Health policies should target regional inequalities as well as socioeconomic inequalities.

Barriers for better emergency medical services in Armenia trigger distrust of the general population towards services Nune Truzyan N Truzyan School of Public Health, American University of Armenia, Yerevan, Armenia Contact: [email protected]

Background Emergency Medical Services (EMS) in Armenia performed in hospitals and operates through ambulance. The purpose of this study was to assess accessibility and performance of EMS in Yerevan, Armenia, reveal barriers for better EMS, and provide evidence for further improvements. Methods To comprehensively assess EMS we applied mixed quantitative-qualitative research methods Data was collected through 1) desk analysis, 2) focus group/indepth interviews with emergency physicians, cardiologists, general practitioners, medical and high school students (N = 22), 3) six-month retrospective and 18-month

Downloaded from http://eurpub.oxfordjournals.org/ by guest on January 16, 2017

Does living in rural environment generate inequalities in the Romanian women attitude regarding cervical cancer early detection?

59

60

European Journal of Public Health, Vol. 24, Supplement 2, 2014

The health inequalities of alcohol related hospital admissions Neil Martin N Martin Balance - The North East Alcohol Office, Durham, UK Contact: [email protected]

Aim There is limited research in regards to alcohol and health inequalities at the local level. Through this study, local authorities across England will be able to quantify the differences that alcohol has on deprived groups in their local population. Method Alcohol related hospital admissions (ARHA) data for 2002/03 – 2011/12, held by the Department of Health, has been aggregated by 2010 deprivation deciles for the 326 local authorities across England. A Slope Index of Inequality (SII) was then calculated for age and sex standardised admission rates by using a regression line of best fit applied to the data points. The gradient of this line is the slope index of inequality, the steeper the slope, the higher the SII value and therefore the greater the level of inequality. Results As the overall rate of alcohol related hospital admissions increases, then so does the SII and therefore the health inequality. England has seen a 113% increase in ARHAs over the last ten years whilst simultaneously the SII has increased by 109%. The SII has increased more sharply for women at 117%, compared to 106% for males, even though females have seen a smaller increase in ARHAs generally over the time period. These effects differ from area to area depending on levels of deprivation and population structure. ARHA rates have however seen bigger increases in the lesser deprived groups

across England with a 115% and a 105% increase for the least and most deprived deciles respectively. This effect is amplified further in the North East with a 135% and 113% increase for the least and most deprived deciles respectively. Conclusion Since staring this research the Department of Health have issued a new methodology to calculate ARHAs. Work is now in progress to update the SIIs for each of the local authorities. Clearly relationships between alcohol and deprivation are complex and further analysis will look in more depth at different types of hospital admission as well as age and gender trends. An advanced understanding of alcohol related health inequalities will hopefully steer the public health agenda in a direction that will ultimately benefit the people who suffer the greatest health harms through the misuse of alcohol. Key messages  The England population disproportionately suffers health inequalities when looking at alcohol related hospital admissions.  Data is available to analyse and understand the complexities surrounding alcohol related health inequalities and this needs to be used by local authorities to develop public health policy.

A health system approach to improve NCD outcomes and reduce inequalities in the Republic of Moldova Alessandra Ferrario M Skarphedinsdottir1, B Smith2, A Ferrario3, O Zues4, A Ciobanu5, M Tirdeau6, S Domente5, J Habicht5, M Jakab7 1 World Health Organization Regional Office for Europe, Copenhagen, Denmark 2 Abt Associates, USAID Quality Project, Bishkek, Kyrgyzstan 3 London School of Economics and Political Science, London, UK 4 WHO Barcelona Office for Health System Strengthening, Barcelona, Spain 5 World Health Organization Country Office of the Republic of Moldova, Chisinau, Republic of Moldova 6 Ministry of Health of the Republic of Moldova, Chisinau, Republic of Moldova 7 WHO Barcelona Office for Health System Strengthening, Barcelona, Spain Contact: [email protected]

Background Several policy documents call for a comprehensive health system response to reduce the burden of non-communicable diseases (NCDs). However, it is unclear what this health system response exactly constitutes in practice. The aim of this study was to explore health system barriers that prevent scaling up of core NCD services and lead to poor and inequitable health outcomes as well as identify good practices and pragmatic policy recommendations that could be implemented in the Republic of Moldova. This is part of a multicountry study led by WHO Regional Office for Europe. Methods The country assessment used a structured assessment guide developed by WHO Euro. This is an innovative tool providing a framework to assess health system bottlenecks in addressing NCDs. Data collection was based on a desk review, fieldwork in Moldova in May 2013 and a final workshop with national stakeholders to validate the findings and the recommendations in February 2014. Results The burden of NCDs in Moldova is unevenly distributed, with mortality from major NCDs showing intracountry geographical and gender disparities as well as intercountry disparities within the region. Coverage of population and individual level interventions to prevent and control NCDs ranged from limited to moderate. For example, evidence suggests that clinicians do not routinely base primary prevention decisions on CVD risk scores, which results in overtreatment with aspirin and undertreatment with statins. Population health coverage has been strengthened in recent years. Since 2010, there is universal coverage of primary health care and emergency services. Payment mechanisms are in place

Downloaded from http://eurpub.oxfordjournals.org/ by guest on January 16, 2017

prospective ambulatory cardiac records review (N = 85), and 4) general population baseline/midterm/follow-up (N = 1140) telephone surveys in 2005-2010. Qualitative data was analyzed using conventional inductive methods Quantitative data analyses included descriptive and analytic Methods Results Considering that in 96% of cases, people in Yerevan called ambulance before collapse happened, we expected high survival rate. However, we disclosed four main categories of barriers for better outcome in cardiac arrests: 1) physical (delays in ambulance arrivals due to poor condition of cars, roads and lack of traffic regulations); 2) technical (insufficient medication and equipment); 3) human (lack of cardiopulmonary-resuscitation (CPR) skills among general population); and 4) social (out-of-pocket payments). Though, the attitude of general population toward EMS still rather negative in 2010, there were systematic pattern of positive change from 2005 on believe that in a case of need the ambulance will respond in a timely manner and that ambulance dispatcher will not request payment (p = 0.04). These caused systematic increase in percent of people who in a medical emergency preferred waiting for the ambulance versus taking patient directly to the hospital (p = 0.01). However, people were significantly less intended to call ambulance in 2010 than in 2005 when their family member (p = 0.002) or a stranger (p = 0.05) had a serious medical emergency. Conclusions Barriers existing in EMS negatively impact CPR outcome and cause distrust of the general population towards emergency services. Better regulations for physical, technical, human, and financial problems are needed in Armenia. Key messages  Cardiac arrests survival rates are low in Armenia due to physical, technical, human, and social barriers.  Insufficient quality of emergency medical services in Armenia leads to population distrust toward the services.

7th European Public Health Conference: Thursday 20 November 2014 16:00-17:00

that improve equity in resource allocation and provide incentives for health care providers to deliver better services. However, although good progress has been made towards more equal access to care, household expenditure on health remains comparatively high; in 2012, health was the fourthhighest household expenditure item at 6%. Conclusions Despite decreasing, overall mortality rates from NCDs in the Republic of Moldova are still comparatively high compared with other countries in the region. Major gains can be made through relatively low-cost public health and primary health care interventions. Key messages  A health-system research design enabled to identify key hurdles and facilitators to improve treatment and outcomes for NCDs and to formulate pragmatic and actionable recommendations.  These findings will serve as evidence-base for national stakeholders to influence and complement national NCD control efforts and will contribute to experience sharing at international level.

Stefanie Rezansoff SN Rezansoff, JM Somers, A Moniruzzaman Faculty of Health Sciences, Simon Fraser University, Burnaby, Canada Contact: [email protected]

Background Problem-solving courts (PSC) have been implemented internationally, with a common objective to prevent reoffending by strengthening social determinants of health and public safety. PSCs are distinguished from traditional adjudication by frequent judicial interaction with court participants, collaborative treatment by an interdisciplinary team, a nonadversarial approach by defence and prosecution, and a strong emphasis on rehabilitation. The Downtown Community Court (DCC) of Vancouver is a PSC that integrates the health, housing, and social resources of the surrounding community in order to promote rehabilitation among offenders with complex needs including substance dependence, mental disorders, homelessness, and poverty. Method We used the propensity score matching method to examine the effectiveness of Vancouver’s DCC. We focused on the subset of participants identified as having the most complex level of need, and who were assigned to a case management team (CMT). A comparison group was derived using one-to-one matching on a large array variables, including sociodemographic, health-related and criminogenic factors, geography, and time. Reductions in offences (one year pre minus one year post) were compared between CMT and comparison groups. Results Compared to other DCC offenders, those triaged to CMT (n = 249) had significantly higher levels of healthcare, social service use, and justice system involvement over the ten years prior to the index offence. Compared to matched offenders who received traditional court outcomes, those assigned to CMT exhibited significantly greater reductions in overall offending (p < 0.001), primarily due to significant reductions in property offences (p < 0.001). Conclusions Our findings indicate that CMT achieved significantly greater reductions in recidivism than traditional court among offenders with complex needs and high numbers of previous offences. Offenders with complex health and social needs may be at increased risk for reconviction unless service gaps such as healthcare and housing are filled. The non-adversarial approach taken by PSCs has been shown to be viable in North America, but may also be beneficial in European

countries where inquisitorial (rather than adversarial) judicial proceedings are normative. Key messages  Compared to traditional adjudication and justice responses, community courts may significantly reduce crime among offenders with complex health and social needs.  The non-adversarial nature of community courts is compatible with the inquisitorial justice systems of most European countries.

Estimating renal replacement therapy (RRT) requirement in Romania: a spreadsheet model which estimates need and addresses inequalities in service provision Andreea Steriu A Steriu1, C Sanderson2 European Public Health Consultant, London, UK 2 London School of Hygiene and Tropical Medicine, London, UK Contact: [email protected] 1

Background Chronic kidney disease stage 5 (CKD5) with end-stage renal failure (ESRF) is uncommon but expensive to treat, despite advances in health technology. Equitable national service provision requires good quality information. Treatment modelling is a mathematical aid to service planning. Methodology and Methods The ‘prevalence = incidence x survival’ equation was used as methodological base. Baseline prevalence was estimated from two data sources with capture-recapture method (CRM); a baseline estimated national acceptance rate was used. One- and 3-year survival probabilities were measured from three centres, in three different regions of the country. Comparison were made with international reporting from: ERA-EDTA, UK and USA. A mathematical spreadsheet model with scenarios was considered to estimate annual numbers of patients with CKD5 in need of RRT. The model was calibrated during 1997-2006. It is validated during 2007-2016. Levels of change for parameters include: unchanged, increased (%) or decreased (%) levels. Acceptance has four levels: baseline + three levels of change, mortality has three levels: baseline + two levels of change (12 output scenarios). Outcomes. Expected number of patients (annual values: numbers and rates pmp). Results The baseline estimate obtained with CRM was 2,995 patients or 175 pmp (rate per million population 15 years +). Baseline acceptance rate was 80 pmp (range 15 to 85 pmp). At the end of the calibration period, results for two out of 12 output scenarios were 6,698 (389 pmp) and 6,978 (406 pmp). These were the closest values to the reported figures of 6,600 (384 pmp) and 7,071 (411 pmp) for the same year. The two output scenarios assumed: 1) unchanged parameters and 2) a 10% increase in acceptance over 10 years with unchanged mortality. Compared with European RRT rates the calibrated prevalence is close to the median European level (437 pmp; range 28 to 579 pmp, ESRD Incidence Study Group). CI95% were calculated for all estimates. Conclusion Spreadsheet modelling is a very useful and flexible service planning tool. It can assist service planning with equitable resource distribution at national level. Local variations need further exploration (weakness of model). Figures need to consider the context of economic affordability and efficiencies. Key messages  Spreadsheet modeling is a flexible planning tool. This model proves that it can be used for service planning for any chronic condition providing there is sufficient good information.  Local variations (inequalities) in service provision can be addressed with estimating need at local level, but uncertainty levels around estimates can increase, especially if conditions are rare.

Downloaded from http://eurpub.oxfordjournals.org/ by guest on January 16, 2017

Examining the impact of case management for offenders with complex health and socio-legal challenges in a Canadian urban context

61

62

European Journal of Public Health, Vol. 24, Supplement 2, 2014

Association between LINE-1 hypomethylation and cancer risk: a systematic review and meta-analysis Martina Barchitta M Barchitta, A Quattrocchi, V Adornetto, A Maugeri, A Agodi Department ‘‘GF Ingrassia’’, University of Catania, Catania, Italy Contact: [email protected]

Downloaded from http://eurpub.oxfordjournals.org/ by guest on January 16, 2017

Background DNA methylation, a major epigenetic mechanism, is involved in various biological processes including cancer. Long Interspersed Nuclear Element-1 (LINE-1) hypomethylation, as biomarker of global methylation, was observed in several types of cancer and was associated with a poor prognosis. In the present systematic review, global DNA methylation was evaluated as a biomarker for cancer risk. Methods A literature search, using the PubMED database, was performed to select studies, published up to March 2014, which associated LINE-1 hypomethylation with cancer. The association between LINE-1 hypomethylation and cancer was described with respect to the study design, the tumour site and the methods used to measure LINE-1 methylation levels. Results The search resulted in the identification of 45 studies published between 2004 and 2014. 80% were case-control studies. A total of 18 studies were from Asian countries (40%), 13 from European countries (29%), and 12 from USA (27%). The most frequent tumor type in study was colorectal cancer (8 studies)

and hepatocellular carcinoma (6 studies). Regarding the method of LINE-1 methylation levels evaluation, ‘‘gold standard method’’, used in most studies (67%), included DNA bisulfite treatment, PCR and pyrosequencing. In 41 studies LINE-1 methylation levels were evaluated both in tumor and in healthy controls tissues. Overall, the studies reported the results obtained from 21 297 samples: 14 251 from cancer patients and 7046 from healthy patients. In 15 studies blood samples were used. Eight studies evaluated LINE-1 methylation levels in patient cohorts with cancer identifying LINE-1 hypomethylation as a biomarker of poor prognosis in cancer patients. Considering 15 studies, meta-analysis shows the lowest mean DNA methylation level in cancer patients compared to the control group (mean difference: 7.21; 95% CI 9.77, 4.47). Conclusions Global DNA hypomethylation appears to be an appropriate biomarker for cancer risk and prognosis. Results suggest that study design, methods used to quantify global DNA methylation levels and cancer type are important factors to be considered in the assessment of the association between hypomethylation levels and cancer risk. Key messages  LINE-1 DNA hypomethylation appears to be an appropriate biomarker for cancer risk and prognosis.  Methods used to quantify global DNA methylation levels are important factors in the assessment of the association between hypomethylation levels and cancer risk.