STRENGTHS AND LIMITATIONS OF THIS STUDY

This sampling strategy ensured the representativeness of the population being studied.

This study was able to gain access to the formal and informal refugee and migrant population residing in Pakistan.

None of the eligible participants refused to be part of the study or withdrew from the study.

The geographic area of study is limited to Pakistan therefore the results cannot be generalised.

Vaccination data were collected based on recall due to the unavailability of vaccination cards.

Introduction

Pakistan is a low-middle-income country and the sixth largest populous nation in the world with over 247 million people,1 2 with a per capita income of US$4.9 per day.3 It is estimated that 38.3% of the Pakistani population is living in multidimensional poverty marked by several factors mainly poor health, lack of education, low income, poor quality of work and threat of violence.4 Approximately, 26.8% of the Pakistani population resides in rural areas where health, hygiene and sanitation conditions are suboptimal compared with urban areas.5

Pakistan is considered the world’s second-largest refugee-hosting country due to its close borders with Afghanistan. Pakistan currently hosts approximately 1.37 million registered people of concern, of which 1.32 million refugees are from Afghanistan.6 But the actual numbers are higher, as over 1.5 million refugees are residing undocumented and these are mainly residing in the provinces of Khyber Pakhtunkhwa and Balochistan, both of which border the war-torn country of Afghanistan. Pakistan also hosts 11 340 registered refugees from Somalia, Burma, Palestine, Iran and Iraq.6 The majority of this population prefers to live in rural areas due to low living costs and also, in most cases, the camps arranged by the government for the living of refugees lie on the outskirts of urban areas. Pakistan also has a large population of migrants from Bangladesh who have been living for more than four decades now.

Pakistan and COVID-19

Pakistan recorded its first case of COVID-19 in its largest metropolitan city, Karachi, causing widespread hysteria in the population, with denial of the existing pandemic. The number of COVID-19 cases exponentially increased, initially owing largely to the pilgrims returning from Iran and inadequate testing facilities, thereby revealing the country’s vulnerability to disasters. Pakistan recorded its first death due to COVID-19 on 18 March 2020 and 45 days later, from the appearance of its first COVID-19 case, 4601 cases of COVID-19 had been recorded, with 66 confirmed deaths.7 8 Stringent lockdowns were implemented throughout the country, and quarantine and isolation centres were established to stabilise the overburdened hospitals. In early April 2020, the country was testing about 5000 people daily,9 which increased to 0.27 per thousand tests per day by February 2022.10

According to the WHO, Pakistan has reported 1 580 631 confirmed cases of COVID-19 with 30 656 deaths between February 2020 and June 2023.11 Since the first COVID-19 case in the country, Pakistan has tackled five different waves over the course of 2 years. While the country’s efforts to curb COVID-19 remain laudable, there were constraints in maximising the country’s potential to sustainably recover from the pandemic.

Pakistan initiated a COVID-19 vaccination drive in February 2021.12 It faced multiple challenges in vaccinating its citizens against COVID-19. It also remained at a disadvantage with the refugees, when in the initial months of the COVID-19 vaccination drive, refugees in the country were unable to receive vaccination due to their unregistered status.13 This resulted in a major hindrance in the country’s effort to promote equality and prevent prejudice among different communities. Refugees and migrants have been on political priority across the world due to forced displacements, yet there is a gap in evidence about both the needs and determinants of COVID-19 vaccination and coverage among refugees, migrants in regular situations (MIRS) or migrants in irregular situations (MIIS) in Pakistan. Therefore, we aimed to produce evidence on the coverage, disparity, barriers and facilitators to COVID-19 vaccination for refugees, MIRS or MIIS in Pakistan impacted by large-scale migration.

Objectives

The objectives of this cross-sectional study were to (1) estimate the proportion of refugees and migrants reached by COVID-19 vaccination campaigns in Pakistan and inequity compared with the general population and (2) understand the COVID-19 vaccine access, uptake, knowledge, enablers and challenges for refugees and migrants in Pakistan.

Methodology

This study was led by the University of Geneva which is a part of a larger study conducted in six different countries spanning different regions of the world: Ecuador (Latin America), Nepal (South Asia), Pakistan (South Asia), the Philippines (South-East Asia), Rwanda (sub-Saharan Africa) and Tajikistan (Central Asia). In this paper, we report the findings from Pakistan only, following the Strengthening the Reporting of Observational Studies in Epidemiology checklist of reporting results (see online supplemental annex 1).14 15 Like in the other five countries, we conducted a mixed-method study comprising of a quantitative study (a cross-sectional study) and a qualitative component (in-depth interviews and focus group discussions), but in this paper, we present the findings from the quantitative study only. The overall study methodology is graphically represented in figure 1.

Figure 1

The overall study framework and methodology.

Quantitative assessment

In the cross-sectional survey, we included refugees, MIRS and MIIS (aged <18 years and above) (see box 1) who qualified for the COVID-19 vaccine in Pakistan. To ensure accurate identification of participants falling under the categories of refugees, MIRS and MIIS, we contacted the government district team which was further supported by the experience of the research team. Additionally, some participants did self-report their refugee and migration status.

Box 1 Operational definitions

Refugee—Refugees are persons who are outside their country of origin for reasons of feared persecution, conflict, generalised violence or other circumstances that have seriously disturbed public order and, as a result, require international protection. The refugee definition can be found in the 1951 Convention and regional refugee instruments, as well as the United Nations High Commissioner for Refugees’ Statute.

Migrant—An umbrella term, not defined under international law, reflecting the common lay understanding of a person who moves away from his or her place of usual residence, whether within a country or across an international border, temporarily or permanently, and for a variety of reasons. The term includes a number of well-defined legal categories of people, such as migrant workers; persons whose particular types of movements are legally defined, such as smuggled migrants; as well as those whose status or means of movement are not specifically defined under international law, such as international students.

Regular migration—Migration that occurs in compliance with the laws of the country of origin, transit and destination.

Irregular migration—Movement of persons that takes place outside the laws, regulations or international agreements governing the entry into or exit from the state of origin, transit or destination.

Only those participants who agreed to participate and provided informed consent were included. The survey was conducted in Quetta (a city in the province of Balochistan), Karachi and Hyderabad (cities in Sindh province) where we included Afghani and Bengali refugees and migrants and these locations were purposively selected as they all had high registered numbers of refugees and migrants. We did not exclude participants based on gender, race, religion, ethnicity or other characteristics. We did plan to exclude participants who lacked the capacity to respond to the survey questions because of any illness or disability; however, we did not come across any participant with illness or disability.

Sampling

We identified Afghan and Bengali refugees and migrants from the areas recommended by the government, including Karachi (Bengali Para), Hyderabad (Afghan Basti) and Quetta (Hazara Town). We adopted the WHO’s 30×7, two-stage cluster sampling strategy in each of these areas to estimate vaccination coverage.16 In the first stage, we randomly selected 30 clusters (neighbourhoods/census blocks) through a computer-generated list, followed by a random selection of seven interview households within each cluster. In each of these clusters, the team collected data from these households starting from an end in the right direction.

Data collection, management and analysis

A structured questionnaire was used in all six countries (online supplemental annex 2). The study captured data on demographics, migration status and COVID-19 infection including data on risks, infection status, testing, hospitalisation, preventive measures and COVID-19 vaccination, knowledge, access and behaviours in the context of infection and vaccination. The questionnaires were piloted on 10 households to assess the feasibility, accuracy and adequacy of the questionnaire and accordingly revised and finalised. These data were not included in the final analysis.

Data were collected by trained field staff through in-person surveys on paper-based forms. The training of data collectors spanned 4 days, incorporating a dedicated day for hands-on training. The training for data collectors in Karachi and Hyderabad commenced on 7 June 2022, whereas in Quetta, it began on 13 June 2022. Subsequently, data collection activities were initiated on 15 June 2022 and successfully concluded on 30 June 2022, in Karachi and Hyderabad. While, in Quetta, data collection commenced on 20 June 2022 and concluded on 15 July 2022. The collected data were entered into RedCap software.17 The information was checked for accuracy and completeness at different levels by the data managers. Only the research team had access to the server to download and check data quality and missing data were reported. All the data collection forms were coded and kept in locked cupboards. The data in the software were archived and stored in a data repository at the Aga Khan University (AKU) in Karachi. Access to the data repository was limited to data management personnel directly involved in the project through their AKU local area network identification with the level of access depending on the role of the user. Data were replicated to a remote location as a backup. A fail-over/slave server was maintained to ensure the database could be restored in the event of a disaster that resulted in downtime for the primary server.

Outcome measures

The primary outcomes of the study were:

Proportion of refugees vaccinated against COVID-19

Proportion of MIRS vaccinated against COVID-19

Proportion of MIIS vaccinated against COVID-19

COVID-19 Vaccine Equity Index (CVEI) by Pressman et al 18

The COVID-19 vaccine equity gap index among refugees and migrants was calculated by using extant literature on the COVID-19 CVEI by Pressman et al.18 The formula to calculate CVEI is given below:

where the subscript ‘s’ denotes a subgroup (ie, vaccinated refugees and migrants) and the subscript ‘t’ denotes members of the entire (or total) population at large (vaccinated population of Pakistan, ie, 63%).19 The ratio of equals to 1. A ratio of greater than 1 for a particular group denotes less favourable and an unequal treatment and a ratio of less than 1 denotes less equal but a favourable treatment.

Analysis

Initial analysis included examining the frequency distribution of all variables to identify possible errors. Final analyses were performed after data cleaning and satisfactory quality assurance. For continuous variables, means, SDs and ranges, and for categorical variables, frequencies and percentages were computed by using STATA V.16.1 software.20 The χ2 test and Fisher’s exact test were used at a 5% level of significance to determine the significant differences in proportions between categories of variables. Fisher’s exact test was specifically chosen when χ2 assumptions were violated.

Ethical considerations and maintaining confidentiality

The study was reviewed and approved by the AKU Ethical Review Committee and the National Bioethics Committee (see online supplemental annex 3). Confidentiality of all collected data was assigned a high priority at each stage of data handling. Written consent was obtained from all the research participants. It was ensured that only female interviewers took consent and interviewed the female respondents. Individual names and personal information of respondents were kept confidential and personal identifiers were not used in any form of reporting. Datasets were also kept anonymous for analysis. All data files were saved in password-protected files.

Patient and public involvement in research

Participants and their families were central to the study and were involved even before the actual data collection and this helped to motivate the community involvement during and beyond the study.

ResultsDemographics

A total of 570 participants were interviewed, of which 324 were from Sindh (Hyderabad and Karachi) and 246 from Balochistan (Quetta) (online supplemental table 1). 51.6% of the respondents were females and 48.4% were males, 94.7% of the participants have been residing in Pakistan for more than 5 years, 64.9% were regular migrants with valid documents, 23% were irregular migrants (with no valid documents), and 12.1% were refugees. Only 14.2% had completed their primary education, 10.5% had completed secondary education, and a majority were illiterate (72.8%) (table 1).

Table 1

Sociodemographic status of participants

COVID-19COVID-19 risk factors and preventive measures

The major risk factors for COVID-19 included demographic attributes, with 21.8% (124/570) of the participants above 60 years of age. 9.47% (54/570) of the participants reported having existing heart disease or suffering from hypertension, while other risk factors included diabetes, chronic lung disease, obesity or weak immune system (online supplemental table 2).

Participants reported taking precautionary measures to avoid contacting COVID-19 by washing hands and using disinfectants (70.4%; 383/544), being careful with washing and cleaning clothes and utensils (56.4%; 307/544), using face masks (38.4%; 209/544), observing social distancing (25.6%; 139/544) and avoiding public gatherings (25.2%; 137/544).

COVID-19 infection

Only 4.56% (26/570) of participants reported being infected with COVID-19, the majority of them were regular migrants. Only 26.9% (7/26) of these cases were confirmed through COVID-19 testing and 3.85% (1/26) of them were hospitalised. 73.1% (19/26) of the participants reported having access to COVID-19 vaccination before contracting COVID-19 but were not vaccinated. The major reason for not getting vaccinated was the experience of being rejected because of one’s foreigner status (42.9%) or due to no vaccination process in place in the country (42.9%) (table 2).

Table 2

COVID-19 infection among participants

COVID-19 infection among participants was significantly associated with geographical location and immigrant status and COVID-19 testing was associated with geographical location.

Knowledge and source of information

More than half (56.8%; 324/570) of the participants were unaware of the available COVID-19 vaccine brands but 56.7% (323/570) were aware of the vaccination site/setting before getting infected with COVID. The major source of information regarding COVID-19 vaccination was from local newspaper/radio/TV channels (58.6%; 334/570), foreign newspapers (28.8%; 164/570) and family or social networks (24.4%; 139/570) (online supplemental table 3).

Willingness, empowerment and trust

About 47.4% of the participants were very concerned about the serious side effects associated with COVID-19. Despite that, only 11.9% (68/570) of the participants were not willing to get vaccinated, 74.4% (424/570) preferred getting vaccinated from a hospital and 23.2% (132/570) from a health centre/clinic.

More than half (53.2%; 303/570) of the participants reported the willingness of their family and friends to get the COVID-19 vaccine but 51.6% (294/570) showed very little trust in healthcare providers. Based on decision-making, 63% (359/570) of the participants were empowered enough to decide whether to get vaccinated or not, although in a few cases, the decision to get vaccinated was taken by a partner/spouse (13.9%; 79/570) or by the father (11.6%; 66/570) (online supplemental table 4).

Willingness and decision-making power to get the COVID-19 vaccine were significantly associated with geographical location and immigrant status. Concerns regarding the side effects of the COVID-19 vaccine and trust in healthcare professionals were associated with geographical location.

COVID-19 vaccination, funding and certification

The government of Pakistan was reported to be the major funder of COVID-19 vaccination (online supplemental table 5). 66% (376/570) of the participants were offered COVID-19 vaccination, and 55% (314/570) received one dose of COVID-19 vaccine. Among those who were not offered or don’t know if they were offered (34%; 194/570) COVID-19 vaccination, only 34% (66/194) of them planned to get vaccinated against COVID-19. Of those offered, 66% (248/376) were fully vaccinated (fully vaccinated was defined as receiving the all required doses of a certain vaccine according to the recommendations at that time and this included single-dose vaccines), 17.6% (66/376) were partially vaccinated, and 14.4% (54/376) were not vaccinated. Very few participants in Balochistan (37.8%; 93/246) were offered COVID-19 vaccine as compared with participants in Sindh (87.3%; 283/324). 63.4% (199/314) of the vaccinated participants had received at least two doses and 32.8% (103/314) had received at least one dose of COVID-19 vaccine. Most of the participants received their first and second doses between January and June of the year 2022 (online supplemental tables 6 and 7). Sinovac and Pfizer were mostly administered as the first and second doses of the COVID-19 vaccine; however, in Balochistan, most of the participants received CanSino (35.9%; 23/64) for their first dose and Sinovac for their second dose (26.3%; 5/19). Only 3.82% (12/314) of the participants received a booster dose of the COVID-19 vaccine, and this was lower among refugees and migrant population residing in Balochistan. 83.3% (10/12) of the participants received their booster dose in the year 2022 and most of them received Pfizer, Moderna and CanSino. Most participants reported not paying for the COVID-19 vaccination.

Of the vaccinated participants, 83.4% (262/314) had a vaccination card/certificate, of which 79.8% (209/262) got it from a vaccination centre and 91.6% (240/262) of them did not face any problem acquiring it. Few faced system issues while acquiring the vaccination certificate (6.87%; 18/262). Most of the participants (62.98%; 359/570) were unaware of the accessibility of the implemented vaccination plan for the people in mobility in their residing country as compared with the neighbouring countries. This was also found to be significantly associated with geographical location and immigrant status.

Being offered with COVID-19 vaccine, COVID-19 vaccination status (ie, fully or partially vaccinated), and having a vaccination card or a certificate was found to be associated with both geographical location and immigrant status. The number of doses of COVID-19 vaccine and the place of obtaining the vaccination card were associated with geographical location and the plan of getting vaccinated without being offered was associated with the immigrant status.

Refusal and side effects

Of those who did not plan to get vaccinated or were unsure (65.9%; 128/194), the major reason included adverse reactions to COVID-19 vaccine (50%; 64/128), unavailability of identification documents (8.59%; 11/128) and fear of syringes/needles and doctors (8.59%; 11/128). Other reasons included permission issues, lack of knowledge and trust (table 3). The reason for getting vaccinated is given in figure 2. Fever was reported as the only side effect of the COVID-19 vaccine.

Table 3

COVID-19 vaccination refusals among refugees and migrants

Figure 2

Reason for getting vaccinated.

Facilitators and barriers to the vaccination process

The difficulties faced by the participants during the vaccination process included distance to vaccination centres (12.3%; 70/570), physical limitations (10.5%; 60/570), unavailability of COVID-19 vaccine for migrants (8.7%; 50/570), difficult registration process (8.7%; 50/570), unavailability of smartphone/internet (6.14%; 35/570), long waiting hours (5.61%; 32/570) and rejections due to unavailability of identification documents (5.44%; 31/570). Participants (15.4%; 88/570) also reported being discriminated during the vaccination process which was mostly observed by healthcare staff (59.1%; 52/88) or the security personnel (33%; 29/88) (online supplemental table 8).

The unavailability of identity documents and perceived differences in treatment for vaccination compared with the local population were associated with geographical location and immigrant status.

Despite several challenges, the participants also reported several facilitators to the vaccination process, which included the generation of a special code by the government to increase vaccination among people who did not have valid identification documents (44.4%; 8/18), provision of timely information on vaccination process (34.6%; 197/570) and development of multiple vaccination centres in different areas of the city to increase accessibility (30.7%; 175/570) (online supplemental table 9).

National healthcare setups were accessible to more than half of the population in Sindh (58.6%; 190/324) as compared with Balochistan (17.5%; 43/246), which was approximately a 30- to 60-minute drive away. Vaccination centres were reported to be hygienic (51.8%; 295/570) and participants received precise and clear messages regarding the COVID-19 situation (48.2%; 275/570) and adequate information on vaccination rates (45.4%; 259/570). Government campaigning was also successful in grabbing the attention of migrants and refugees to go to vaccination centres in Sindh (63%; 204/324); however, migrants and refugees in Balochistan were mostly unaware of it (64.6%; 159/246). All these factors were found to be significantly associated with geographical location and immigrant status.

All the descriptive tables and the stratified analysis by education, age, gender, immigration status, years of living and country of origin are given in online supplemental tables 10–17.

COVID-19 Vaccine Equity Index

A CVEI of 0.87 (ie, the ratio of less than 1) was calculated using the Pressman et al 18 literature and the results show a less equitable but favourable treatment with the refugee and migrant population in Pakistan. The calculations are given below:

Discussion

The study highlights COVID-19 vaccine coverage, inequity, access, risk factors, knowledge, willingness, challenges and enablers in accessing COVID-19 vaccines by refugees, MIRS and MIRS residing in Pakistan. The study surveyed a total of 570 participants of which 65% were regular migrants with valid documents, 23% were irregular migrants (with no valid documents) and 12.1% were refugees. The majority of the participants were illiterate. The major risk factor among them was older age (>60 years) and the existence of a comorbidity. The majority of the participants did report on taking precautionary measures to avoid contracting COVID-19. Despite these measures, few contracted coronavirus (4.56%; 26/570) and were hospitalised due to COVID-19 (3.85%; 1/26). Among them were very few who got themselves tested for COVID-19 (26.9%; 7/26). COVID-19 infection among participants was found to be significantly associated with geographical location and immigrant status, while COVID-19 confirmation through tests was found to be associated with geographical location.

About 66% (248/376) of the refugees and migrants reported themselves as fully vaccinated, and 17.6% (66/376) as partially vaccinated, which was also significantly associated with geographical location and immigrant status. The most common reason for not getting vaccinated was being rejected due to foreign nationality. Vaccination rates varied across provinces, genders and different migrant populations. The percentage of the vaccinated population was low in Balochistan compared with Sindh and low among irregular migrants and refugees compared with the regular migrant population. Sinovac and Pfizer were mostly administered as the first and second doses of the COVID-19 vaccine, while Pfizer, Moderna and CanSino were mostly administered as a booster vaccine. Fever was reported as the only side effect of the COVID-19 vaccine. When questioned on willingness to get vaccinated, 88% (502/570) showed a willingness to get vaccinated and preferred to get vaccinated from a hospital or a health clinic. Most refugees and migrants were empowered enough to make their own decision regarding COVID-19 vaccination but for a few the decision was dependent mostly on the male member of the family (mostly be it a father or a spouse).

Despite the vaccine campaigns by the government, 14.4% (54/376) of the refugee and migrant population remained unvaccinated mostly because of the vaccine not being offered, distant vaccination sites, access, unavailability of COVID-19 vaccine or due to difficult registration process. Interestingly, the government being the major funder of the COVID-19 vaccine, still few had to pay a fee to get an appointment for the vaccination.

In terms of displacement status, low- and middle-income countries host 85% of the refugee population.21 Unfortunately, the distribution of COVID-19 vaccinations has been markedly uneven on a global scale, with wealthier nations receiving a disproportionate supply compared with developing countries.22 23 This gap widened with each new wave and new variant.21 This study estimated a CVEI of 0.87 showing less equitable but favourable access to COVID-19 vaccination within the migrant and refugee populations. Similar challenges were observed in various countries, including Pakistan, where vaccine inequity and hesitancy were prevalent reasons for the low COVID-19 vaccine coverage among refugee and migrant populations.23–25 According to the reports, only 35% of the refugees in Morocco were fully vaccinated compared with 60% of the nationals.24 A similar pattern was observed in Angola (9% refugees vs 11.9% nationals), South Africa (10% refugees vs 26% nationals), Egypt (10.3% refugees vs 20% nationals), India (23% refugees vs 72% nationals) and Uganda (<1% refugees vs 8% nationals).21 24

Consistent with the findings of our study, studies have reported on the willingness of refugee and migrant populations to receive the COVID-19 vaccine, as well as the barriers to accessing it.26 These studies have highlighted challenges such as poor access to vaccination due to lack of identification documents and discrimination against refugees, asylum seekers and migrants residing in countries such as Uganda, Iran and India, and in high-income countries like the USA, Greece and the UK.21–23 25 Misconceptions and information gaps were reported in Uganda, where information was disseminated in local languages rather than those spoken by the refugee populations.22 To overcome these barriers, some countries, akin to Pakistan, lifted the requirement of identification documents to increase vaccination coverage.22 Some countries collaborated with non-governmental organisations (NGOs) and other health networks to improve the dissemination of health information and reduce vaccine hesitancy among the migrants and refugee populations.22

With joint efforts of government, NGOs and development partners, Pakistan has somehow maintained a balance between economic hardships and mortalities due to disease. It is now at a critical junction of the pandemic where it has witnessed a substantial decrease in the number of cases but cannot afford complacency.

Limitations of the study

The study’s findings cannot be generalised to high-income countries as it is more focused on refugees and migrants residing in the context of Pakistan. Second, verification of vaccine status was a major limitation as the vaccination card was not available for all at the time of the survey and the data were collected based on recall.

Suggestions for future policies

All policies should prioritise inclusivity for all refugees and migrants. Documentation of refugee and migrant populations should be prioritised, not only for future emergencies but for daily life and equal access to facilities. In future, the registration mechanism for vaccinations should be much simpler than the one that was initially established in Pakistan. The primary aim should be to curb the rapid spread of disease and to avoid the valuable loss of time in complex registration processes which were identified as a bottleneck. Law enforcement agencies and departments concerned with refugees and migrants should be involved in policymaking, as they play a major role in identifying and locating vulnerable communities. All communication should be carried out in the local language of the area. The local community should be engaged and mobilised so that there is greater success in achieving results.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statementsPatient consent for publication

Not applicable.

Ethics approval

ERC approval was taken from the National Bioethics Committee (NBC) (ref: No 4-87/COVID-107/22/73) and Aga Khan University (ref: 2022-7336-21669). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We would like to acknowledge the support of the community members and government officials.