1 Access to non-communicable disease health care among non-camp Syrian refugees in northern JordanManuela Rehr1, Muhammad Shoaib1, Idriss Ait-Bouziad2, Cono Ariti3, Paul van den Bosch1, Sara Ellithy1, Suhib Okour1, Anais Deprade1, Mohammad Altarawneh4, Abdel Alshafee4, Sadeq Gabashneh4, Annick Lenglet5 1 Médecins Sans Frontières Jordan, 2 Médecins Sans Frontières UK, 3London School of Hygiene & Tropical Medicine, UK 4 Ministry of Health Jordan, 5 Médecins Sans Frontières Netherlands

2 Background & objectivesPopulation & health care needs UNHCR has registered >600,000 Syrian refugees in Jordan (78% outside refugee camps) Key health care needs: non-communicable diseases (NCDs) MSF treats >3500 refugees with NCDs in northern Jordan (OCA/OCBA) Access to health care Since 11/2014, non-camp refugees pay for health care (subsidized rate) Previous surveys suggested limited access to care including NCD care Objectives Determine access to health services for non-camp Syrian refugees (general adult & child health, NCDs, ANC) Health care needs, service utilization & expenditures Barriers to health care

3 Methods Study design Target population & definitionsCross-sectional household survey, two-stage cluster: 329 randomly selected clusters over inhabited areas x 8 households by snowball sampling Target population & definitions Non-camp Syrian refugees in Irbid governorate Households arriving after Dec. 2011 NCD section: only adults ≥ 18 years Self-reported chronic conditions: hypertension, diabetes type I/II, cardiovascular conditions, thyroid disease, chronic respiratory conditions, cancer

4 Methods & survey overviewProcedures Cluster starting points: random GPS coordinates Mobile data collection (ODK) 18 data collector teams Survey overview Data collection: 22nd May to 28th June 2016 2589/2712 households consented (95.5%) 17,854 individuals covered (12.4% of all non-camp Syrian refugees)

5 Socio-demographics & -economics45.7% were ≥ 18 years (8,041/17,854) 55.7% Education 44.3% 12.6% (95%CI: ) no formal education 35.5% 28.1% 60.8% (95%CI: ) at least secondary education 14.4% 11.7% 5.8% 4.5% 18-39 y 40-59 y >60 y (N=8,041); cell proportions (N=8,041)

6 Socio-demographics & -economicsAverage monthly household income & expenditures Average household size 6.8 (95%CI: ) 79.3% (95%CI: ) of households in debt (N=2,589) Income 337 US$ Expenditures 506 US$ In the month preceding the interview, (income: N=2194), (expenditures N=2287). Income included UNHCR-cash-assistance.

7 Adult NCD prevalence (self-reported)Hypertension 14.0% (95%CI: 13.3 – 14.8) Diabetes (type I/II) 9.2% (95%CI: 8.5 – 9.9) Cardiovascular conditions 5.7% (95%CI: 5.1 – 6.2) Chronic respiratory diseases 3.2% (95%CI: 2.8 – 3.6) Thyroid disease 2.5% (95%CI: 2.2 – 2.9) Cancer 0.6% (95%CI: 0.4 – 0.7) (N=8,041). In persons ≥ 18 years of age.

8 of adults suffer from at least one NCDMorbidities & co-morbidities 44.7% reported more than one chronic condition Cardiovascular diseases only 5.8% Hypertension & CVD 7.1% Hypertension, diabetes & CVD 8.1% Respiratory disease only 9.1% Diabetes only 10.1% Hypertension & diabetes 17.6% Hypertension only 22.7% (N=1,756). Other combinations not shown (19.5%) 21.8% of adults suffer from at least one NCD (95%CI: ) (N=8,041)

9 did not seek NCD medical care the last time it was neededAccess to NCD care One randomly selected NCD patient per household (N=1,243) Reasons 75.4% (95%CI: ) had at least one MD consultation 61.5% Affordability 12.7% Knowledge (N=806). In the 6 months prior to the survey. 9.6% Availability 22.9% (95%CI: 20.4 – 25.5) did not seek NCD medical care the last time it was needed 5.4% Approachability 0.8% Acceptability (N=1,133). (N=260). Other reasons (7.7%) & unknown (2.3%) not shown.

10 Seeking NCD care: Risk factor analysisDesign-based multivariate logistic regression (N=941) Adjusted for: age, gender, education, type of NCDs, household income & -size Seeking NCD care when needed Unadjusted OR (95%CI) Adjusted OR (95%CI) Wald test n/N 18-39 years 127/195 .reference 40-59 years 343/446 1.78 ( ) 1.16 ( ) ≥60 years 277/325 3.09 ( ) 1.96 ( ) Age p=0.040 Type of NCD CVD 195/260 ) 0.70 ( ) Any other NCD but CVD 552/706 .reference Diabetes 376/434 2.81 ( ) 2.40 ( ) Any other NCD but diabetes 371/532 Hypertension 514/626 2.11 ( ) 1.81 ( ) Any other NCD but hypertension 233/340 p=0.057 p<0.001 p=0.001

11 Seeking NCD care: Risk factor analysisContinued Seeking NCD care when needed Unadjusted OR (95%CI) Adjusted OR (95%CI) Wald test n/N Poorest (1st quintile) 174/234 .reference (2nd quintile) 207/276 1.03 ( ) 0.99 ( ) (3rd quintile) 89/116 1.14 ( ) 1.01 ( ) (4th quintile) 139/182 1.11 ( ) 1.04 ( ) Richest (5th quintile) 138/158 2.38 ( ) 2.11 ( ) Household income p=0.083 Little evidence for an association: Gender, education, other NCDs, household size

12 Health care utilization & expendituresSector utilization 6.9% of mean monthly household income Public Private NGO = 23.0 US$ mean costs/visit (range ) 27.1% 18.0% 51.1% (N=873). Others & unknown not shown (3.8%). (N=783). Per consultation including provider fees, medication and diagnostics; Includes all patients that received free-of-charge services.

13 had a medication interruption for >2weeksAccess to NCD medication 92.2% needed regular medication for their NCD Reasons for medication interruption 63.4% Affordability 23.1% (95%CI: 20.3 – 25.9) had a medication interruption for >2weeks in the past 6 months 9.1% Symptoms improved & patient stopped 6.0% Did not know where to get medication 4.9% MD instructions 9.8% Others Among NCD patients who need regular medication (N=1,146). (N=265). Unknown reasons (6.8%) not shown.

14 Limitations Self-reported NCDs  potential under- or over-reportingProlonged limited access to health care  under-diagnosis of NCDs & real prevalence could be higher Snowball sampling  clustering; sample size assumed a design effect of 2.0 (results btw. 1.0 – 1.4) Economics: focussed on income only. Further work required to determine the “ability to pay”.

15 Summary & conclusions Health care needsNCD is a key requirement: one in five adults suffers from at least one NCD High caseload from hypertension and diabetes Access to NCD care Majority did access care but approx. a quarter of patients did not seek care when needed &/or experienced NCD medication interruption Barriers to care Reported unaffordability of provider costs High income patients more likely to seek NCD care High financial household burden for NCD medical care visits Even for those who did access care: opportunity costs? Increased support for NCD health services for Syrian refugees are required by engagement of national, regional & international stakeholders

16 Acknowledgements UNHCR in Jordan. Sundus A. Bbdullah and Hazim Q. Rawashdh from ACTED, Jordan. Ayat Nashwan and Abdel B. Athamneh from the Yarmouk University, Ali Oudat from the Jordanian Red Crescent/German Red Cross . Directorates of Health in Irbid and Ramtha. MSF team members Kiran Jobanputra, Erwin Guillergan, Luis Eguiluz, Anne Garella, Rachida Aouameur, Claudia Nicoletti, Maya Abu Ata , Luna Hammad, Fatima O. Mohamed, Omer A. Abdalla, Tawfeeq Abu Agha, Mohammad Al-Faris, Atwa Bakeer, Jebreel Barham, Marjan Besuijen, Christopher E. Boehm, Rowa S. Ennab, Juniper Gordon, Tariq Hamad, Imee J. Japitana, Motaz Omari, Jamal Qwaider and Sameer Rawashdeh. The data collectors & team drivers

17 Crude NCD prevalence International Diabetes Foundation (IDF)MENA (2014): 9.7% Crude Jordan (2014): 8.9% Crude Syria (2014): 7.4% Crude Syria (2011): 8.2% NCD-RisC Group HTN (Lancet) Crude Jordan (2014): 18.9% (men), 13.8% (women) Crude Syria (2014): 21.3% (men), 19.4% (women) NCD-RisC Group Diabetes (Lancet) Crude Jordan (2015): 13.1% Crude Syria (2015): 11.9% Doocy et al (for UNHCR, 2014) HTN Among Syrian refugees in Jordan: 10.7% Doocy et al (for UNHCR, 2014) DM Among Syrian refugees in Jordan: 6.1%