Skip to main content
NCBI home page
Sage Choice logoLink to Sage Choice
. 2021 Apr 14;33(8):861–869. doi: 10.1177/10105395211000913

Private Health Insurance in Malaysia: Who Is Left Behind?

Nur Zahirah Balqis-Ali 1,✉, Jailani Anis-Syakira 1, Weng Hong Fun 1, Sondi Sararaks 1
PMCID: PMC8592113  PMID: 33853361

Abstract

Despite various efforts introduced, private health insurance coverage is still low in Malaysia. The objective of this article is to find the factors associated with not having a private health insurance in Malaysia. We analyze data involving 19 959 respondents from the 2015 National Health Morbidity Survey. In this article, we describe the prevalence of not having health insurance and conducted binary logistic regression to identify determinants of uninsured status. A total of 56.6% of the study population was uninsured. After adjusting for other variables, the likelihood of being uninsured was higher among those aged 50 years and above, females, Malay/other Bumiputra ethnicities, rural, government/semigovernment, self-employed, unpaid workers and retirees, unemployed, lower education level, without home ownership and single/widowed/divorced, daily smoker, underweight body mass index, and current drinker. The likelihood of being uninsured also increased with increasing household size while the inversed trend was seen for household income. A substantial proportion of population in Malaysia did not have private health insurance, and these subgroups have limited preferential choices for provider, facility, and care.

Keywords: Malaysia, National Health and Morbidity Survey, private health insurance, uninsured, vulnerable group

What We Already Know

  • Malaysia health system is dichotomous, with almost free of charge public health care and fee for service private health care.

  • Private health insurance in Malaysia serves as a secondary role of duplicating the public health services.

  • Private health insurance allows population greater access to private health care that offers preferential access to timely care, choice of provider and care options, as well as reduced waiting time.

What This Article Adds

  • The population more likely to be uninsured and had limited preferential access to health care were vulnerable groups including those aged 50 years and above, females, Malay/other Bumiputra ethnicity, rural, government/semigovernment, self-employed, unpaid workers and retirees, unemployed, lower education level, without home ownership and single/widowed/divorced, daily smoker, underweight body mass index, and current drinker. The likelihood of being uninsured increased with increasing household size while the inversed trend was seen for household income.

  • These vulnerable groups need to tap into financial resources such as savings to enable private inpatient care or continue accessing the public health care, which is under pressure from congestion and concern of the sustainability from current financing method.

  • Despite various initiatives introduced to reduce the burden of these vulnerable groups in getting adequate health services, more effort needs to be in place by the government to increase their preferential access to health care.

Introduction

In countries with large publicly funded health systems, private health insurance (PHI) fortifies the system by increasing access to both public and private health sectors.1,2 In these settings, PHI serves a secondary role of either supplementing, complementing, or duplicating public health services, which may come at the expense of access for public patients. 3 Through these roles, PHI allows greater access to private health care that offers preferential choice for timely care, provider, care options, and waiting time, which may contribute toward health and wellbeing. 3

In Malaysia, private hospitals establishment began in 1950s and progressively increased to 210 hospitals in 2019.4,5 Corporatization of certain public health care facilities, like the National Heart Institute and the availability of preferential services, provided more options for treatment with varying charges.4,6 There were obvious disparities between the services offered by the public and private sectors, with shorter waiting times and the ability to choose your own doctor deemed better served in the private health sector. 7 Twenty-one percent of the population reported waiting times to see doctor in public hospitals as inadequate, compared with 3.2% for private hospitals. 7 Similarly, choice of health care provider was rated inadequate (14.9%) by respondents who received public inpatient care, compared with 3.0% for private hospitals.4,7 However, for the population without PHI coverage, payment option includes paying out-of-pocket (OOP). This results from a dichotomous health care system of heavily tax-financed public health care accessible almost free of charge (a maximum of Malaysian Ringgit [MYR] 500/US$127.98 is incurred per hospitalization for third-class public-sector ward charges), 8 and a private sector that relies mostly on OOP financing, allowing the more affluent better access to private sector health care. 4

Although Malaysia has a pro-poor public and a pro-rich private sector with a government spending of slightly more than half of the 4.3% of GDP on total health expenditure (THE), equality of access and financial risk protection is good. 9 Despite recognized achievements in universal health coverage, sustainability remains a challenge. 6 The country’s percentage for OOP per THE showed an increasing trend, from 29.6% in 2000 to 34.2% in 2010 and 38% in 2017, 4 reflecting the demand for private health care. In 2017, private sector sources of financing, amounting to 48.9% of THE, were mainly from private household OOP (77%), while private insurance, corporations, and other private sector agencies contributed 15%, 5% and 3%, respectively. 10 The use of private sector reduces congestion and cost pressure through voluntary opt-out from public services. However, the population could return to public health care at any time, which typically occurs due to cost or the need for prolonged care. 11 The PHI in Malaysia is sold under Medical and Health Insurance (MHI) policy and has been in the Malaysian market since 1970s. These policies can be attained on voluntary basis either by individual purchase or through employer-sponsored health insurance (ESHI) scheme. 12 The PHI mostly covers private hospital inpatient treatments, with the consumer usually paying the bill upfront prior to being reimbursed by the insurance company, 13 and can be used in public facilities. 14 As of 2018, there were 210 private hospitals accommodating 15 957 beds (27.3%), mostly in urban areas, as opposed to 144 public hospitals with 42 424 beds nationwide. 5 PHI serves duplicative or additional option for purchasers since utilization of public facilities are not affected. It is risk-rated, whereby increasing age or individuals with preexisting conditions may either be charged higher premiums or precluded from purchase. 13 Additionally, premiums vary greatly depending on extent and duration of coverage. 15

Despite long-standing establishment, the MHI sales in Malaysia only surged in the 1990s after the introduction of personal income tax relief in 1996. The government’s decision to allow stand-alone MHI policies in 1997 also helped catalyze the market’s growth. Most insurers use the service of managed care organizations for more efficient administrative processes as well as curbing the total administrative cost. 12 The MHI market was regulated by the government through the Central Bank of Malaysia, mainly through the guidelines and regulations provided under the Private Health Care Facilities and Services 1998. Efforts were made to encourage the uptake of PHI in Malaysia, including addressing the needs of certain target groups, such as the elderly. 12

In Malaysia, 15% of the population were estimated to have some form of coverage in 2005. 12 In 2015, 23.7% had individual purchase and 15% was covered by ESHI, with some overlaps. 7 Although most of the population in Malaysia do not have PHI, little is known about the characteristics of this group. This article explores factors associated with not having PHI using data from a community survey in 2015. This provides insights to government and policymakers on subgroups who have limited access to preferential care.

Methods

We used data on adults aged 18 years and above from the National Health and Morbidity Survey (NHMS) 2015, a nationally representative community survey. 7 Respondents were classified as uninsured when they did not have PHI or ESHI. However, they might have guarantee letter, which is public sector coverage for civil servants. Independent variables included characteristics of the individual, household, as well as behavioral and chronic illness status.

Proportions with 95% confidence intervals (CIs) were used to describe uninsured rates and binary logistic regression (enter method) to identify determinants of uninsured status with survey weights applied in Stata 13, with survey tab and logistic commands for the respective analysis. We analyzed percentage of children per household size instead of number of children and number of household members, as latter two variables resulted in collinearity. The model with number of household members and percentage of children per household size resulted in the best fit model (Pearson’s correlation coefficient = 0.2177), while the Pearson’s correlation coefficient between household members and number of children was 0.6934 (highly correlated). Adjustment for household cluster effect was tested during model development and dropped as there was no differences observed. Odds ratio (OR) and 95% CI of OR were presented for regression analysis. All costs were expressed in MYR and US$ using the average exchange rate for 2015, US$1 equaled to MYR3.907. 16

Results

In total, 56.6% (95% CI = 54.9, 58.2) was uninsured (Table 1). Differences were seen across age, sex, ethnicity, education, occupation, marital status, geographical strata, household income, household size, home ownership, and chronic illnesses, with higher proportions of uninsured among those aged 60 years and above, female, Malay/other indigenous ethnicity, poorest two quintiles, self-employed, unpaid or retiree, primary or no formal education, and single/widowed/divorced. In contrast, around half of the Chinese, private employee, and richest two quintiles have higher rates of PHI.

Table 1.

Private Health Insurance (PHI) Coverage.

Variables Total Uninsured Insured
PHI a ESHI b Total insured
Count Estimated population % 95% CI % 95% CI % 95% CI % 95% CI
, 19 959 20 238 623 56.6 54.9, 58.2 26.8 25.3, 28.4 16.6 15.4, 17.9 43.3 41.8, 45.1
Sex
 Male 9497 10 449 571 53.1 51.1, 55.0 26.8 25.1, 28.6 20.12 18.4, 22.0 46.9 45.0, 48.9
 Female 10 462 9 789 052 60.3 58.5, 62.1 26.8 25.2, 28.6 12.9 11.8, 14.0 39.7 37.9, 41.5
Ethnicity
 Malay 12 349 9 918 392 65.8 64, 67.6 18.5 17.2, 20 15.6 14.5, 16.8 37.8 36.0, 39.5
 Chinese 3203 4 759 245 42.1 39, 45.4 52.2 49, 55.4 5.7 4.7, 6.9 59.5 56.3, 62.6
 Indian 1410 1 371 927 50.8 46.2, 55.4 35 30.4, 40 14.2 11.7, 17 52.5 47.9, 57.0
 Other Bumiputra 1759 2 178 975 73.4 68.9, 77.4 14.7 12, 18 11.9 9, 15.5 31 26.9, 34.4
 Others 1238 2 010 084 62.2 55.6, 68.5 7.7 5.6, 10.7 30 24.1, 36.7 40.8 34.2, 47.7
Strata
 Urban 11 518 15 416 827 54.8 52.8, 56.7 31.7 29.8, 33.7 17.2 15.8, 18.6 48.9 46.9, 50.8
 Rural 8441 4 821 796 77.3 74.8, 79.6 11.2 9.8, 12.8 14.9 12.3, 17.9 26.1 23.4, 29.0
Citizenship
 Malaysian 18 687 18 181 050 56.3 54.7, 57.8 29.0 27.5, 30.6 14.7 13.7, 15.7 43.7 42.2, 45.3
 Non-Malaysian 1272 2 057 573 59.1 52.3, 65.6 7.5 5.3, 10.4 33.4 27.3, 40.1 40.9 34.5, 47.7
Work status
 Government/semi-government 2194 1 924 753 37.1 34, 40.3 44.8 41.6, 48 18.1 15.8, 20.7 62.9 59.7, 66
 Private 6110 8 013 108 37.8 35.5, 40.2 31.3 28.9, 33.7 31 28.7, 33.3 62.2 59.9, 64.5
 Self-employed 3837 3 402 320 69.4 66.8, 71.9 25.9 23.5, 28.4 4.7 3.8, 5.8 30.6 28.1, 33.2
 Unpaid worker 3327 2 884 448 75.1 72.8, 77.4 18.1 16.1, 20.3 6.8 5.7, 8.1 24.9 22.6, 27.2
 Retiree 786 609 294 73.9 69.7, 77.8 20.1 16.6, 24.2 5.9 4, 8.8 26.1 22.2, 30.3
 Unemployed 3686 3 377 829 80 77.4, 82.4 15.9 13.7, 18.3 4.1 3.2, 5.2 20 17.6, 22.6
 Missing 19 26 871 93 60.8, 99.1 7 0.9, 39.2 0 — 7 8.7, 39.2
Education
 No formal 1370 1 179 253 81.5 76.7, 85.5 7 5.3, 9.1 11.5 8.0, 16.4 18.5 14.5, 23.3
 Primary 4786 4 107 419 74.8 72.2, 77.3 12.2 10.7, 14.0 12.9 10.8, 15.4 25.2 22.7, 27.8
 Secondary 9150 9 349 073 57.8 55.9, 59.68 24.7 23.1, 26.4 17.5 16.1, 19.0 42.2 40.3, 44.1
 Tertiary 4398 5 280 862 34.3 32.1, 36.5 47.2 44.7, 49.8 18.5 16.9, 20.2 65.8 63.5, 67.9
 Unclassified 216 277 800 60.8 49.7, 70.8 12.9 7.2, 22.1 26.3 17.1, 38.3 39.3 29.2, 50.3
 Missing 39 44 216 73.9 50.0, 88.9 11.1 3.3, 31.2 15 5.0, 36.9 26.1 11.1, 50.0
Marital status
 Never married 4194 5 645 620 54.3 51.7, 56.8 25.2 23.2, 27.3 14.7 13.2, 16.2 45.7 43.2, 48.32
 Married 13 811 13 165 992 55.2 53.4, 57.0 28 26.3, 29.7 16.7 15.3, 18.1 44.8 43.0, 46.6
 Widow/widower/divorcee 1940 1 404 857 78.1 75.3, 80.7 13.7 11.5, 16.3 8.1 6.5, 10.1 21.9 19.3, 24.7
 Missing 14 22 154 100 — — — — — 0 —
Age group, years
 18-29 4909 6 602 553 54.8 52.3, 57.2 25.3 23.1, 27.5 20 18.1, 22 45.2 42.8, 47.7
 30-39 4011 4 711 981 46.3 43.7, 48.9 32.8 30.4, 35.3 20.9 18.8, 23.2 53.7 51.1, 56.3
 40-49 3666 3 535 319 49 46.4, 51.7 34.1 31.5, 36.9 16.8 15, 18.8 51 48.3, 53.6
 50-59 3579 2 777 119 61.9 59.3, 64.5 25.3 23, 27.7 12.8 11.2, 14.6 38.1 35.5, 40.7
 60-69 2360 1 635 800 80 77.4, 82.3 14.7 12.6, 17.2 5.3 4.1, 6.8 20 17.7, 22.6
 >70 1434 975 851 91.3 89.0, 93.2 6.7 5.1, 8.8 0.2 1.2, 3.3 8.7 6.8, 11
Household income quintile c
 Q1 2801 2 367 125 80.5 77.5, 83.2 13.6 11.6, 16.0 5.9 4.4, 7.8 19.5 16.8, 22.5
 Q2 3726 3 433 076 74 71.3, 76.4 10.5 9.0, 12.2 15.6 13.3, 18.1 26.1 23.6, 28.7
 Q3 4306 4 178 479 63 60.3, 65.6 18.7 16.5, 21.0 18.3 16.2, 20.7 37 34.4, 39.7
 Q4 4108 4 422 161 51.5 48.7, 54.3 28.7 26.3, 31.3 19.8 17.7, 22.0 48.5 45.7, 51.3
 Q5 5018 5 837 782 35.9 33.5, 38.3 46.2 43.5, 48.9 17.9 15.9, 20.2 64.1 61.7, 66.5
Number of household members
 1 1442 1 551 302 51.9 47.8, 55.9 29.5 25.8, 33.4 18.7 15.4, 22.5 48.1 44.1, 52.2
 2 3550 3 327 297 58.2 55.1, 61.2 26.4 23.8, 29.1 15.4 13.3, 17.8 41.8 38.9, 44.9
 3 3638 3 646 727 56.6 53.7, 59.4 26.5 23.9, 29.2 16.9 15.0, 19.1 43.3 40.6, 46.3
 4 3892 4 102 102 53.1 50.2, 55.9 30.4 27.6, 33.3 16.6 14.7, 18.7 46.9 44.1, 49.8
 5 2930 3 086 994 56.4 53.1, 59.6 27.9 24.8, 31.2 15.7 13.5, 18.2 43.6 40.4, 46.9
 6 2022 2 011 063 57.5 58.0, 66.9 24.8 21.6, 28.3 17.7 14.9, 20.9 42.5 39.0, 46.1
 ≥7 2485 2 513 138 62.6 58.0, 66.9 20.8 17.5, 24.6 16.7 13.3, 20.7 37.5 33.1, 42.1
Percentage of children in household
 <50% 16 545 16 855 997 57.5 55.8, 59.3 26.3 24.7, 28.0 16.2 14.9, 17.6 42.5 40.7, 44.2
 >50% 3414 3 382 626 51.8 48.9, 54.7 29.6 26.9, 32.4 18.6 16.4, 21.0 48.2 45.3, 51.1
Home ownership
 Owned 15 083 14 361 218 58 56.3, 59.8 29.7 28.0, 31.5 12.3 11.4, 13.2 42 40.3, 43.7
 Staff quarter 1182 1 000 706 42.2 34.9, 49.8 19 14.0, 25.3 38.8 30.6, 47.7 57.8 50.2, 65.1
 Rented 2849 3 942 160 56.6 53.2, 59.9 20.5 18.0, 23.3 22.9 20.2, 25.8 43.4 40.1, 46.9
 Others 614 694 475 49.1 38.6, 59.6 12.2 8.2, 17.6 38.8 28.3, 50.4 50.9 40.4, 61.4
 Missing 231 240 064 52.1 42.3, 61.7 31.8 23.6, 41.4 16.1 10.6, 23.7 47.9 38.2, 57.7
Body mass index
 Underweight 1107 1 259 036 63.2 59.1, 67 24.9 21.4, 28.8 11.9 9.6, 14.7 36.8 33, 40.9
 Normal 7796 8 553 140 55.7 53.6, 57.9 26.7 24.8, 28.7 17.6 15.8, 19.5 44.3 42.1, 46.4
 Overweight 5949 5 634 849 55.8 53.5, 58.1 27.8 25.9, 29.9 16.4 14.8, 18.1 44.2 41.9, 46.5
 Obese 3645 3 313410 57.6 55.1, 60.1 25.2 22.9, 27.5 17.2 15.5, 19.1 42.4 39.9, 44.9
 Missing 1462 1 478 188 56.6 52.1, 60.6 29.1 25.1, 33.4 14.5 12.0, 17.5 43.4 41.8, 45.1
Drinking status
 Nondrinker 17 905 17 243 244 58.7 57.1, 60.2 16.9 23, 25.9 24.5 15.7, 18.2 41.3 39.8, 42.9
 Ex-drinker 226 309 938 55.2 45.8, 64.3 30.4 227, 39.4 14.3 8.4, 23.4 44.8 35.7, 54.2
 Current drinker 1064 1 688 560 36.8 32.3, 41.5 48.4 43.5, 53.3 14.8 12.2, 18 63.2 58.5, 67.7
 Unclassified 715 934 526 51.9 44.9, 58.9 32 26.7, 38 16 10.3, 24.1 48.1 41.1, 55.1
 Missing 49 62 355 87.4 64.5, 96.4 1.9 0.3, 11.8 10.7 3.3, 29.3 12.6 3.7, 35.5
Smoking status
 Daily smoker 3970 4 389 034 57.6 55.0, 60.2 19.7 18, 21.6 22.6 20.3, 25.1 42.4 39.8, 45
 Occasional smoker 379 457 595 53.3 46.5, 64.0 26.4 20.8, 32.9 20.2 15.5, 26.1 46.7 40, 53.5
 Former smoker 472 493 579 55.8 49.7, 61.6 24.5 19.9, 29.7 19.8 15.5, 24.9 44.2 38.4, 50.3
 Never smoked 15 104 14 853 046 56.3 54.6, 58 29.1 27.4, 30.9 14.6 13.5, 15.9 43.7 42, 45.4
 Missing 34 45 369 97 80.7, 99.6 0 — 3 3.9, 19.3 3 0.4, 19.3
Physical activity
 Active 13 290 13 358 360 55.8 54, 57.6 26.5 24.9, 28.1 17.7 16.3, 19.3 44.2 42.4, 46
 Inactive 6409 6 596 977 57.9 55.7, 60 27.9 25.8, 30 14.3 12.9, 15.8 42.1 40, 44.3
 Missing 260 283 286 56.6 54.9, 58.2 20.7 14.5, 28.7 17.2 11.7, 24.4 37.9 29.5, 47.0
Hypertension
 No 12 711 14 088 532 52.7 50.9, 54.5 29.3 27.5, 31.2 18 16.6, 19.5 47.3 45.5, 49.1
 Yes 7224 6 114 010 65.3 63.3, 67.2 21.2 19.6, 22.9 13.5 12.2, 14.9 34.7 32.8, 36.7
 Missing 24 36 081 96.3 76.1, 99.5 0 — 3.8 0.5, 23.9 3.8 0.5, 23.9
Diabetes mellitus
 No 15 706 16 672 680 54.6 52.9, 56.4 28 26.4, 29.6 17.4 16, 18.8 45.4 43.6, 47.1
 Yes 4229 3 529 862 65.4 62.8, 67.9 21.7 19.5, 24 13 11.5, 14.7 34.6 32.2, 37.2
 Missing 24 36 081 96.3 76.1, 99.5 0 — 3.8 0.5, 23.9 3.8 0.5, 23.9
Dyslipidemia
 No 9421 10 561 034 55.7 53.7, 57.7 26.9 25.2, 28.7 17.4 15.9, 19 44.3 42.3, 46.3
 Yes 10 514 9 641 508 57.4 55.6, 59.2 26.8 25.1, 28.6 15.8 14.5, 17.2 42.6 40.8, 44.4
 Missing 24 36 081 96.3 76.1, 99.5 0 — 3.8 0.5, 23.9 3.8 0.5, 23.9
Open in a new tab

Abbreviations: ESHI, employer-sponsored health insurance; CI, confidence interval.

a

Overall, 28.1% (95% CI = 25.9-30.5) of those with PHI have ESHI.

b

Excluded are respondents with PHI.

c

Household income was calculated based on total monthly household income and the respondents were divided into quintiles.

In the final model, all the above variables were significant predictors except chronic illnesses (Table 2). Additionally, OR for body mass index and level of physical activity were not significant after adjusting for other variables in the model.

Table 2.

Factors Affecting Uninsured Status.

Variables Crude OR (95% CI) Adjusted OR (95% CI) P
Sex
 Male 1.000 (ref) 1.000 (ref)
 Female 1.343 (1.249, 1.445) 1.194 (1.057, 1.349) .004**
Ethnicity
 Chinese 1.000 (ref) 1.000 (ref)
 Malay 2.419 (2.072, 2.824) 2.546 (2.114, 3.065) <.0001***
 Indian 1.329 (1.072, 1.647) 1.242 (0.958, 1.611) .101
 Other Bumiputra 3.265 (2.584, 4.126) 2.504 (1.963, 3.195) <.0001***
 Others 2.133 (1.579, 2.882) 2.523 (1.835, 3.469) <.0001***
Strata
 Urban 1.000 (ref) 1.000 (ref)
 Rural 2.706 (2.293, 3.193 1.408 (1.192, 1.663) <.0001***
Home ownership
 Owned 1.000 (ref) 1.000 (ref)
 Staff quarters 0.529 (0.386, 0.724) 0.501 (0.338, 0.742) .001**
 Rented 0.942 (0.811, 1.095) 1.31 (1.114, 1.54) .001**
 Others 0.697 (0.456, 1.067) 0.522 (0.339, 0.805) .003**
Work status
 Private employee 1.000 (ref) 1.000 (ref)
 Government/semi-government 0.971 (0.828, 1.138) 1.441 (1.19, 1.746) <.0001***
 Self, employed 3.736 (3.262, 4.279) 3.222 (2.785, 3.728) <.0001***
 Unpaid worker/homemaker 4.974 (4.337, 5.704) 3.431 (2.919, 4.032) <.0001***
 Retiree 4.669 (3.745, 5.821) 2.694 (2.027, 3.581) <.0001***
 Unemployed 6.593 (5.588, 7.778) 3.458 (2.839, 4.212) <.0001***
Education
 Tertiary 1.000 (ref) 1.000 (ref)
 No formal education 8.447 (6.249, 11.418) 2.302 (1.675, 3.164) <.0001***
 Primary 5.709 (4.908, 6.641) 2.672 (2.209, 3.232) <.0001***
 Secondary 2.629 (2.365, 2.924) 1.67 (1.462, 1.909) <.0001***
 Unclassified 2.971 (1.889, 4.673) 1.928 (1.141, 3.256) .014*
Marital status
 Married 1.000 (ref) 1.000 (ref)
 Never married 0.963 (0.87, 1.066) 1.476 (1.261, 1.728) <.0001***
 Widow/widower/divorcee 2.901 (2.457, 3.425) 1.24 (1.004, 1.531) .045*
Age group, years
 30-39 1.000 (ref) 1.000 (ref)
 18-29 1.405 (1.248, 1.581) 1.019 (0.875, 1.186) .813
 40-49 1.117 (0.981, 1.272) 0.95 (0.816, 1.107) .512
 50-59 1.887 (1.652, 2.156) 1.39 (1.152, 1.677) .001**
 60-69 4.637 (3.913, 5.494) 2.391 (1.894, 3.017) <.0001***
 70/max 12.219 (9.295, 16.062) 5.914 (4.054, 8.629) <.0001***
Household income quintile
 Q4 1.000 (ref) 1.000 (ref)
 Q1 3.889 (3.146, 4.808) 2.487 (1.937, 3.192) <.0001***
 Q2 2.671 (2.285, 3.122) 2.503 (2.082, 3.009) <.0001***
 Q3 1.604 (1.396, 1.843) 1.628 (1.4, 1.894) <.0001***
 Q5 0.526 (0.456, 0.607) 0.578 (0.493, 0.678) <.0001***
Number of household members
 1 1.000 (ref) 1.000 (ref)
 2 1.29 (1.068, 1.56) 1.634 (1.299, 2.055) <.0001***
 3 1.21 (0.997, 1.468) 1.977 (1.545, 2.529) <.0001***
 4 1.05 (0.87, 1.268) 2.649 (2.06, 3.408) <.0001***
 5 1.199 (0.978, 1.469) 2.602 (1.973, 3.431) <.0001***
 6 1.254 (1.011, 1.556) 2.815 (2.123, 3.732) <.0001***
 >7 1.55 (1.212, 1.982) 3.821 (2.818, 5.181) <.0001***
Percentage of children in household
 <50% 1.000 (ref) 1.000 (ref)
 ≥50% 0.793 (0.702, 0.897) 0.668 (0.574, 0.779) <.0001***
Chronic illnesses
Diabetes mellitus
 No 1.000 (ref) 1.000 (ref)
 Yes 1.568 (1.401, 1.756) 1.079 (0.939, 1.24) .281
Hypertension
 No 1.000 (ref) 1.000 (ref)
 Yes 1.687 (1.548, 1.837) 1.104 (0.981, 1.242) .101
Hypercholesterolemia
 No 1.000 (ref) 1.000 (ref)
 Yes 1.072 (0.992, 1.159) 0.908 (0.819, 1.007) .068
Smoking status
 Nonsmoker 1.000 (ref) 1.000 (ref)
 Daily smoker 1.057 (0.955, 1.169) 1.268 (1.09, 1.474) .002**
 Occasional smoker 0.888 (0.681, 1.157) 0.987 (0.712, 1.368) .936
 Former smoker 0.98 (0.771, 1.245) 0.945 (0.703, 1.27) .707
Body mass index
 Normal 1.000 (ref) 1.000 (ref)
 Underweight 1.362 (1.143, 1.624) 1.249 (1.017, 1.535) .034*
 Overweight 1.002 (0.906, 1.107) 0.963 (0.853, 1.087) .54
 Obese 1.079 (0.961, 1.212) 1.016 (0.883, 1.168) .825
Alcohol intake status
 Nondrinker 1.000 (ref) 1.000 (ref)
 Ex, drinker 0.868 (0.594, 1.267) 1.262 (0.749, 2.126) .382
 Current drinker 0.409 (0.337, 0.498) 0.67 (0.532, 0.845) .001**
 Unclassified 0.761 (0.576, 1.005) 0.993 (0.754, 1.307) .959
Physical activity status
 Active 1.000 (ref) 1.000 (ref)
 Inactive 1.087 (0.997, 1.185) 1.035 (0.933, 1.148) .52
Open in a new tab

Abbreviations: OR, odds ratio; CI, confidence interval.

*

P < .05. **P < .01. ***P < .001. Pseudo R2: 0.2366.

Those aged 50 years and above, females, Malay/other Bumiputra ethnicity, rural, government/semi-government, self-employed, unpaid workers and retirees, unemployed, lower education level, without home ownership and single/widowed/divorced, daily smoker, underweight body mass index, and current drinker had higher likelihood to be uninsured. The likelihood of being uninsured increased with increasing household size while the inversed trend was seen for household income.

Discussion

In a country without social- or community-rated health insurance, there is still a substantial proportion of the population without PHI. Those aged 50 years and above, females, Malay/other Bumiputra ethnicity, rural, government/semigovernment, self-employed, unpaid workers and retirees, unemployed, lower education level, without home ownership and single/widowed/divorced, daily smoker, underweight body mass index, and current drinker had higher likelihood to be uninsured. The likelihood of being uninsured increased with increasing household size while the inversed trend was seen for household income. Chronic illnesses did not affect uninsured status.

Although Malaysia introduced measures such as personal income tax relief and stand-alone sale of insurance products policy since 1996, 12 uninsured rate decreased from 85% in 2005 12 to 57% in 2011, 17 and in this study, 56.6%. The likelihood of being uninsured is significantly associated with income and education, similar to China and Singapore.18,19 This finding is also in line with previous study done based on the same survey conducted in 1996, which found lower income and education were significant factors influencing PHI acquisition. 13 Similarly, the report produced by the Central Bank of Malaysia in 2017 showed the uptake of insurance and takaful was low among population from bottom 40% of household income category (B40) as compared with the national population. 20 The median household income was MYR5873 (US$1503), and MYR3166 (US$810) for B40 in 2019. 21 The 2019 household expenditure survey showed that most expenditures were spent on housing, water, electricity, gas and other fuels, food, restaurants, and transportations (68.3%). Only 2.1% was spent on health. 21 With the PHI premiums set to be at least MYR200 (US$51) yearly and continues to increase rapidly, 22 lower income group is less willing and able to pay premiums. This may be explained through the “loss aversion” theory whereby purchasing an additional and nonmandatory item such as a PHI is perceived as a greater monetary loss than the benefit it may offer. 23 Furthermore, public health care in Malaysia is widely available and accessible. In response to this, the government launched PeKa B40 in 2019, a health subsidy initiative aiming to cover some of health expenditures borne by the B40 including health screenings, payment for medical equipment, cancer treatments, and transportation. 24 Another scheme, mySalam was also introduced, offering population with lower incomes a onetime payment for critical illness claim and MYR50/day (US$13/day) of hospitalization benefit, for a maximum 14 days per annum. 25 Both are not equivalent to the PHI and the preferential access benefit that it offers.

Level of education is directly related to the capacity to accumulate and understand health-related information in making decision to purchase health insurance. 26 Lower level of education may restrict understanding or overwhelm an individual, and this could lead to “omission bias” whereby one prefers status quo than making a hard decision. 23 Lower education is also directly related to income, affecting the ability to purchase insurance.

This study showed that those self-employed, unemployed, and retired had higher likelihood of being uninsured compared with private sector employees. This could be partly explained by the fact that many private employers provide ESHI incentives for their workers. However, this is conditional on continued employment.

Among the major ethnic groups in Malaysia, Malay, other Bumiputra, and other ethnic groups had higher likelihood of being uninsured, compared with the Chinese. An analysis of 2006 NHMS data found that race-religion influences individual health insurance demand in Malaysia, 27 similarly in Singapore. 19 Although not the subject of this study and may require further exploration, a possible explanation was the income gap between ethnicities in Malaysia. In 2016, the mean monthly household incomes for Chinese, MYR8750 (US$2240), was 1.4 times higher than Malay and other Bumiputra MYR6267 (US$1604). 21

In this study, those aged 50 and above were more likely to be uninsured, a finding observed in other countries, in which the likelihood of being uninsured increases with age.28,29 Lower insurance coverage among older age group may be related to the relative young market of PHI in Malaysia, higher premium, risk stratification, and age limit that increased from 60 to 70 years in 2005. 12 This inadvertently leaves the older with less choices, limiting them to public health care if they are unable to afford private services.

Household size and composition affect insurance status. The household composition of adult to child ratio, with lower proportion of children, larger households, and females were more likely to be uninsured. In China, Singapore, and England, females were also less likely to be insured.18,19,29,30 As the study assessed the coverage of PHI as individuals and not as a family with cross-coverage, there is a possibility of underestimation, as someone could already covered by a family-based health insurance but was unaware of this and not recorded.

Those who live in rural areas were significantly more likely to be uninsured than those living in urban areas, similar to China. 18 A likely contributing factor is private hospital distribution that tend to be clustered in urban areas, thus a disincentive to those living in rural area to own PHI. 23 The study showed there was no significant association between being uninsured and chronic illnesses, suggesting “adverse selection” may not contribute substantially to PHI purchase in Malaysia. A study using NHMS 1996 and 2006 data found similar findings. 13

As the health insurance is risk rated, these vulnerable subgroups have limited health care options and may likely resort to public facilities or have to pay OOP for private care. They would need to tap into financial resources such as savings to enable private inpatient care, with the public sector a crucial safety net as the OOP for private is higher than public (see the appendix). Without PHI, majority of the population utilizes public health sector, causing more crowding and congestion in the already pressured public health facilities.4,6 Coupled with increasing health expenditures and concerns on the sustainability of continuous subsidy on public health services, the government is considering more sustainable health financing models for the future. 6

It is well known that when health insurance is not properly regulated, it may exacerbate inequalities, allowing only those who can afford or those with low risk to have coverage and preferential access to private health care. This increases the risk of cream skimming whereby insurance provider has the ability to “choose” customers who will increase their profits and cost escalation for the country. 31 It is thus important to look at the pattern and factors associated with not having PHI to understand the barriers and difficulties in widening the coverage.

Between 2016 and 2019, MHI claims grew by 11.6% a year. 32 With the introduction of the Private Health Care Facilities and Services Act 1998, all PHI suppliers must be registered with the managed care organizations for monitoring. The guidelines were revised in 2005 to increase policyholders’ protection including introduction of mandatory minimum free look period and reduced waiting time for claims. 12 There is a plan to further increase the transparency of the various costs covered and adopting co-payments or deductibles to allow policyholders better control on the costs of their health care decisions. 32

The use of a large nationally representative sample in this survey provides community-level estimation of health insurance status for the individual. We used survey set function to adjust for the survey design. However, this survey may underestimate individuals with family health insurance if the individual is not aware of the coverage. This could explain why household clustering effect was not observed in the analysis. The cross-sectional nature of this study did not allow temporal effect analysis on certain variables such as chronic illnesses. Although the factors included in the survey were extensive, it did not cover factors specifically associated with health insurance coverage such as knowledge on health insurance, size effect of insurance premiums, self-assessment of health, and exemption from payment within universal health coverage. Moreover, this study did not explore the effect of PHI on health care utilization and impact toward the Malaysian health system landscape, perhaps areas for future study.

Conclusion

A substantial proportion of population in Malaysia did not have PHI, and vulnerable subgroups were more likely to be uninsured. Therefore, access to private health care is constrained, which translates into loss of opportunities toward having preferential access to timely care, choice of provider and care options, and reduced waiting time.

Acknowledgments

We thank the Director-General of Health, Ministry of Health Malaysia, for permission to publish this article, and Institute for Public Health for data provision.

Appendix

Comparison of Out-of-Pocket (OOP) Payment for Public and Private Inpatient Carea.

OOP Public Private
Median (Q1-Q3) MYR 80 (36-200) 3000 (1200-6500)
US$ 20 (9-51) 768 (307-1664)
Open in a new tab

Abbreviation: MYR, Malaysian Ringgit.

a

 The median OOP reported for private is 37.5 times higher than public inpatient care. This analysis uses data from the National Health and Morbidity Survey 2015, excluding zero OOP payment. 17

Footnotes

Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.

Authors’ Note: The dataset analyzed for this article is part of the National Health and Morbidity Survey 2015 study. Request for data can be obtained from Dr Mohd Azahadi Omar (drazahadi@moh.gov.my), Head of Sector for Biostatistics & Data Repository, National Institutes of Health, Ministry of Health Malaysia, and with the permission from the Director-General of Health, Malaysia.

Ethics Approval: The NHMS survey protocol was approved by the Medical Research and Ethics Committee (MREC), Ministry of Health Malaysia (NMRR-14-1064-21877).

Informed Consent: All participants provided written informed consent prior to interviews during the National Health and Morbidity Survey 2015’s data collection.

ORCID iD: Nur Zahirah Balqis-Ali Inline graphic https://orcid.org/0000-0002-1590-5457

References

Articles from Asia-Pacific Journal of Public Health are provided here courtesy of SAGE Publications

RESOURCES