Comparison of the Number and Reasons for Self-Perceived Barriers to Accessing Primary Health Care Services Between Roma and Ethnic Albanians

The study population consisted of all participants who reported barriers to accessing PHC services (N = 533), which was pertinent to a population-based survey that included a nationwide representative sample of adults of both sexes who attended PHC services during October 2024. The sample for this survey was allocated across 18 sites (districts) in all four regions of Albania (northern, central, southern, and Tirana – the capital of Albania) using probability proportional to size (PPS). Within each selected health facility (n = 18; one per site), participants were recruited consecutively. Accordingly, the approach employed was designed to ensure nationwide representativeness of our study sample.

The minimum required sample size was calculated for several hypotheses related to sociodemographic correlates of self-perceived barriers to accessing PHC services, including gender and socioeconomic characteristics. The significance level (two-tailed) was set at 5%, and the study power at 80%. Based on conservative assumptions, the minimum required sample size was estimated at approximately 800 individuals, stratified by the four Albanian regions and weighted by their respective population sizes. We decided to target 1200 individuals to increase the study power and account for potential nonresponse. Of the initially targeted sample, 168 individuals were either too sick or refused to participate.

Hence, overall, the sample included in this survey consisted of 1032 adult users of PHC services (53% women; mean age: 47±18 years; overall response rate: 86%). Of these, 533 (≈52%) individuals reported barriers to accessing PHC services, thereby constituting the sample of the current analysis.

Data collection consisted of an interviewer-administered semi-structured questionnaire, derived from validated instruments used in several previous studies conducted in Albania. Overall, six interviewers were involved in data collection, having been trained to employ standardised procedures, including consistent terminology and methodological approaches.

The questionnaire inquired about perceived barriers to accessing PHC services, as well as the sociodemographic characteristics and health profiles of study participants.

Initially, all participants included in this survey (N = 1032) were asked whether they had ever faced any barriers (obstacles) to accessing PHC services for themselves or their close family members. Potential responses were: “yes” vs. “no”.

Participants who responded positively constituted the current study population (N = 533) and were subsequently asked the following question: “What barriers do you face in accessing PHC services? (select all that apply): Distance to health facility; Cost of services; Lack of information; Poor service quality; Long waiting times; Transportation issues; Language barriers; Cultural barriers; Health reasons (too sick to go); No trust in the health system; Stigma; Don't know where to go; Other reasons (please specify in your own words)”.

Health profile included assessment of self-rated health of participants (trichotomised into: very poor/poor, average, and good/excellent), presence of any disability (“yes” vs. “no”), and presence of any chronic diseases/conditions (“yes” vs. “no”).

Sociodemographic characteristics included sex (men vs. women), age (in the analysis trichotomised into: 18–39, 40–59, and ≥ 60 years), place of residence (in the analysis dichotomised into: urban and/or periurban vs. rural areas), ethnicity (dichotomised into: ethnic Albanian vs. other ethnicities), marital status (dichotomised in the analysis into: “married” vs. “other”), educational attainment (trichotomised into: low, middle, and high), and economic situation (good, average, and bad).

The study was approved by the Ethical Council of the University of Medicine, Tirana (no. 2507/1, date: 27. 09. 2024). All participants were initially informed about the aim and procedures of the study and assured of the anonymity and confidentiality of the study. In addition, all participants were informed that the findings would be presented in aggregate form, thereby precluding the identification of personal information.

All individuals provided oral consent to participate in the study.

Fisher's exact test was used to assess differences in the distribution of background characteristics (sociodemographic factors and health profile) (Table 1), as well as differences in the distribution of the number and perceived reasons for barriers to accessing PHC services (Table 2) between ethnic Albanians and other ethnicities. Conversely, a general linear model and a binary logistic regression were employed to assess the association between the number of perceived barriers to accessing PHC services and ethnicity groupings (Table 3). Crude and multivariable-adjusted mean values and their respective 95% confidence intervals (95%CIs) and p-values were calculated (upper panel: results from the general linear models), along with odds ratios (ORs: “≥ 2 barriers” vs. “one barrier only” to accessing PHC services) and their respective 95%CIs and p-values (lower panel: results from binary logistic regression). For the binary logistic regression models, we decided to dichotomise the number of barriers into “only one” vs. “≥ 2”, considering the distribution of the number of reported barriers (62 ethnic Albanians reported three or more barriers compared with only 24 ethnic minority participants; Table 2 – lower panel). All multivariable-adjusted logistic regression models met the Hosmer-Lemeshow criterion (17).

P ≤ 0.05 was considered as statistically significant for all statistical tests. Statistical Package for the Social Sciences (SPSS, version 19.0) was employed for all the statistical analyses.

Our study involved a nationwide representative sample of PHC users who reported barriers to accessing PHC services. Minority groups exhibited greater social vulnerability, with markedly higher levels of low educational attainment, economic hardship, poor health status, disability, and chronic conditions. Overall, the cost of services, a key structural barrier, was the most commonly perceived obstacle across all ethnic groups. Ethnic Albanians reported more frequently a lack of trust in the system and dissatisfaction related to expectations, indicating greater systemic barriers in this group. In contrast, minority participants (almost entirely pertinent to Roma and Egyptian communities) reported significantly higher levels of cultural and language barriers, underscoring communication-related obstacles and the absence of multilingual services in transitional Albania. Furthermore, minority groups faced a greater cumulative burden of access barriers, with a higher proportion reporting multiple challenges. Regression analyses revealed that minorities had significantly increased odds of encountering two or more barriers, even after adjusting for a range of sociodemographic and health-related factors. Nevertheless, given the cross-sectional nature of the study, the associations identified between ethnicity and reported barriers cannot establish causality but only highlight correlations that may be influenced by unmeasured contextual factors.

In our study, the cost of health services was the most prevalent barrier to accessing PHC services in ethnic Albanians and Roma/Egyptian minorities. The latter is in line with a previous report from countries of Central and Eastern Europe, including Albania, where the main reason for unmet health needs among Roma minorities consisted of affordability of care (4). Hence, the cost of services is the main reason for the remarkably high prevalence of unmet needs among Roma minorities (4). In Albania, Roma individuals have been reported more than twice as likely to experience unmet healthcare needs compared to their non-Roma counterparts, mainly due to service cost (4, 18). Likewise, mainly due to affordability issues, the odds of experiencing unmet healthcare needs have been reported significantly higher among Roma minorities compared with non-Roma groups in: Bosnia and Herzegovina (4, 18); Bulgaria (4, 18, 19); Montenegro (4, 18); North Macedonia (4, 18); Romania (4, 18, 20, 21); and Serbia (4, 18).

In principle, PHC users in Albania are not required to pay for services. However, medications prescribed by family physicians that fall within the reimbursement list are often perceived as ineffective, prompting patients to incur substantial out-of-pocket expenses for alternative medicines not covered by reimbursement. In addition, informal payments may represent another significant financial barrier to accessing PHC services in Albania.

Our findings related to a higher prevalence of communication-related barriers to accessing PHC services, including language and cultural barriers among Roma/Egyptian minorities, are generally compatible with previous reports from Albania (15, 16). Conversely, our study did not reveal a significant difference between stigma and minority status, as documented in previous studies conducted in Albania (15, 16). The lack of relationship in our study may reflect underlying collinearity, in which stigma is indirectly embedded within more salient, structurally mediated dimensions of exclusion, such as language and cultural barriers.

Our findings related to Roma/Egyptian minorities also resonate with other marginalised groups, such as migrants, refugees, and other population groups in fragile settings, which remain underserved and face significant barriers to accessing healthcare services (12, 22). In particular, language barriers are a major obstacle to migrant healthcare access, and therefore translation and cultural mediation are essential to overcoming these barriers and ensuring equitable care (12).

On the other hand, in our study, ethnic Albanians perceived more systemic barriers than Roma/Egyptian minorities, including a lack of trust in the public system and expectations-related dissatisfaction. These perceived barriers may reflect greater system awareness among ethnic Albanians, who, unlike minorities, may be more accustomed to navigating public services and thus more likely to appraise systemic weaknesses critically. Also, it may reflect greater reliance on private healthcare services among ethnic Albanians, whose higher expectations may predispose them to perceive the public sector more negatively and criticise its shortcomings, as suggested by previous studies in Albania (23, 24).

Our findings related to adult Roma/Egyptian users of PHC services are also compatible with recent evidence from Albania pertaining to children aged 12–15 years (25, 26). In this age-group, Roma/Egyptian children reported: a significantly higher prevalence of barriers to accessing healthcare services (51% vs. 42% among ethnic Albanians), with distance and non-availability of the parents as main barriers (25); and a higher prevalence of poor attitudes toward health promotion (ability to maintain and improve health: 33% vs. 29%, respectively) (26). Furthermore, among Albanian children aged 11–15 years, the prevalence of unhealthy dietary habits (including breakfast skipping and low consumption of fruits and vegetables) has been reported to be lower among lower socioeconomic groups, a population segment that includes Roma/Egyptian minorities (27).

In our study, ethnic minorities reported markedly higher barriers related to culture and language, indicating communication-related obstacles which reflect deeper inequities in the organisation of health services in post-communist Albania. Hence, there is an urgent need for culturally sensitive and language-accessible services in Albania (15). From this perspective, targeted, culturally sensitive, linguistically accessible, and community-based educational programs are needed to increase health literacy levels and system navigation skills among ethnic minorities in Albania and other European countries (15). Furthermore, to address the documented access inequities, policy responses in Albania should prioritise the development of multilingual PHC staff training programs, the systematic inclusion of Roma mediators in community health outreach, and the incorporation of equity monitoring indicators into PHC performance evaluation frameworks. In turn, ethnic Albanian participants were seemingly more affected by systemic barriers, suggesting the need for reforms in service delivery. In particular, the significantly higher expectations than those met by the current public PHC services among ethnic Albanians highlight the need for policy measures that enhance the responsiveness and quality of public PHC services, particularly by aligning service delivery with the expectations of more system-engaged populations such as ethnic Albanians, while ensuring equity across all groups (28, 29).

Conversely, our finding that the cost of health services was the most commonly perceived structural barrier to accessing PHC services across all ethnic groups highlights the urgent need for policy interventions that reduce financial barriers to PHC (such as expanding coverage, subsidising essential services, or introducing targeted fee exemptions) to ensure equitable access across all ethnic groups (30).

Overall, our findings reflect broad challenges in the Albanian health system's responsiveness and the persistent inequities shaping access to PHC in a transitional post-communist health system. The concentration of structural, cultural, and communicative barriers among Roma and Egyptian minorities illustrates a pattern of differential responsiveness, in which services fail to adequately adapt to the social, linguistic, and cultural needs of marginalised communities, thereby undermining equitable access (31). Conversely, the prominence of cost-related barriers across all ethnic groups points to inadequate financial protection in Albania, which is an essential component of a responsive and equitable health system (30, 31). In turn, the greater distrust and expectations-related dissatisfaction reported by ethnic Albanians highlight how unmet expectations also signal weaknesses in responsiveness at the interface of service quality, provider-patient interactions, and system accountability (30, 31). Together, these patterns suggest that barriers to PHC in Albania are not merely individual experiences but manifestations of systemic asymmetries in how the health system accommodates diverse populations, reinforcing the need for reforms that strengthen equity, cultural competence, and the capacity of services to respond appropriately to PHC users' expectations and needs (30, 31).

The cross-sectional design restricts causal inference; while associations between ethnicity and perceived barriers were identified, temporal relationships cannot be established. Additionally, the data collection process was restricted to October 2024. Extending the study duration to three months could have increased participation and improved the representativeness of the sample. Also, reliance on self-reported data may introduce recall and social desirability biases, particularly in sensitive domains such as trust and stigma. In particular, social desirability bias may have led minority participants to underreport barriers they perceive as socially undesirable (e.g., trust, stigma, or discrimination), potentially attenuating the observed ethnic differences in these important barriers. Conversely, recall bias may have led participants to overemphasise more memorable difficulties (e.g., cost, distance, or transportation), potentially exaggerating the share of these structural barriers across all ethnic groups. On another matter, because participants were asked not only about their own experiences but also about those of close family members, responses may have been subject to recall bias or misrepresentation, as individuals might inaccurately perceive or report barriers others face. Additionally, although the sampling strategy ensured national representation, the relatively small sample size (n = 553) may restrict the generalizability of our findings across diverse population groups. Especially, the small number of minority participants (n = 89), particularly non-Roma and non-Egyptian groups (only two participants), limits subgroup analysis and may obscure intra-group heterogeneity. Overall, the relatively small sample size and limited representation of minority groups may have influenced our results by underrepresenting key subpopulations and reducing statistical power; therefore, our findings should be interpreted as indicative rather than definitive, highlighting trends that warrant further investigation in larger, more balanced samples. Furthermore, the questionnaire did not capture the frequency or severity of barriers, nor did it explore institutional or provider-level factors that may contribute to deeper inequities. Lastly, the absence of qualitative data restricts a deeper understanding of cultural and linguistic mismatches, which were shown to affect minority groups in Albania disproportionately. Hence, there is a need for longitudinal, mixed-methods research to confirm and, eventually, expand the findings of our study.