In 39 days with a total of 24 data collection personnel and three field supervisors, we were able to collect data from 1,096 schools in two towns of Karachi. Fifty-eight (5%) schools refused to provide data to our field teams. The reasons for refusals were: absence of an authorized person to provide data at the time of the visit, uncertainty by school administration about the vaccine type and need for its use at the school level, and previous unsatisfactory experience with vaccination or health-related activity. In total, 304,836 students were enrolled in public, private, and religious schools in Gulshan-e-Iqbal and Jamshed towns. Twenty-five percent (276) of all schools had an enrolment of less than 100 students whereas 3% (36) had an enrolment of more than or equal to 1,000 children (Table ).
Basic characteristics of schools in two townships of Karachi, Pakistan
The average minimum school fee for schools charging a fee ranged from 615 to 821 rupeesb (7.60 to 10.18 USD) per month. Jamshed town had a lower median minimum school fee of 350 rupees (4.30 USD) per month. A welfare system to support the educational expenses of deserving students was available in 56% of public schools, 71% of private schools, and 66% of Madrasahs. Approximately 8% of public schools, 17% of private schools, and 14% of Madrasahs provided some form of health education (Table ).
One-quarter of public schools, 41% of private schools, and 23% of Madrasahs previously participated in a vaccination campaign that used the school as a venue. The majority (275/1,090) of these school-based campaigns were for polio vaccination, which was carried out under Pakistan's polio eradication program. Ninety-five schools provided vaccines that were not included in the government immunization programs (e.g., EPI, polio eradication program, and measles second chance program). These vaccines were typhoid (provided in 5 schools), varicella (3 schools), and hepatitis B (89 schools). Many of these vaccines were sold at a price, especially in the private schools (Table ).
Vaccination experience of schools in two townships of Karachi, Pakistan
Vaccine price was frequently cited (59% of respondents) as first choice as a barrier to vaccination of children by parents in schools in Gulshan-e-Iqbal town. In Jamshed town, the proportion was even higher (71%). Among other factors that were considered to be important barriers to a school-based vaccination campaign in the two towns were vaccine side effects (24% in both Gulshan-e-Iqbal and Jamshed), and lack of awareness about the vaccine (34% in Gulshan-e-Iqbal and 38% in Jamshed). Almost one-third of the respondents did not select a third choice when considering factors that may promote vaccine uptake but among those who did respond, risk of disease, vaccine side effects, and lack of awareness about the vaccine were considered to be the most important factors that could affect the vaccination campaign (Table ).
Factors identified by respondents from school authorities as barriers to a vaccination campaign in schools in two townships in Karachi, Pakistan
The respondents were then asked in a similar fashion to comment on factors that may increase student participation in the immunization program (Table ). In Gulshan-e-Iqbal town, permission from parents was the first choice among 67% of the respondents, followed by distribution of vaccine-related information materials (47%) and involvement of teachers (28%). In Jamshed town, the school administration staff considered distribution of vaccine-related information materials as an important factor (74%), followed by involvement of the City District Government (55%), and involvement of teachers (32%).
Factors identified by respondents for school participation in a school vaccination program in Karachi, Pakistan†
The school census activity in two towns of Karachi provided information on vaccine use and vaccine acceptance, in addition to characteristics (e.g., number of schools, fee structure, and student enrolment) of the schools. Our data from school census shows that there is wide variation in characteristics of the educational institutions of Karachi, Pakistan. Our results also show that majority of education services are provided by private sector. Health and education have recently been dominated by private sector in Pakistan [20
]. The reasons of such transition have been well documented for Pakistan and other developing countries that include quality of service, consumer satisfaction and access. In such a situation, private schools' significant role not only in education but also in health cannot be ignored. Immunization policies in Pakistan therefore should take into account schools' role and private schools in particular to meet global standards of immunization coverage.
Mapping of the schools identified that many schools are located in close proximity to each other on major streets in the study setting. The smaller schools are located in sections of houses of the school owner or administrator. This type of small school may present logistical issues at the time of the vaccination campaign regarding accessibility. The student size at private schools varies from as low as five students at a school to approximately 3,000 students. Such variety in school size also presents a challenge in terms of properly scheduling a vaccination campaign.
Regular health programs offered at the school level were not very common in this study setting. Increasing health awareness in students in schools may be limited due to limited existing resources at the schools. School-based vaccination campaigns may have to adopt strategies that are attractive and innovative in order to create the need and demand for vaccines in schools. This is in line with the global effort to integrate health activities at schools such as FRESH and WASH campaigns [22
]. Furthermore, based on the census teams' interviews with school administration, participation rates were very low in immunization programs where the vaccine was offered at a price.
The majority of schools were willing to participate in a school-based vaccination campaign but were unsure of the response from parents. Though there are indications from other endemic areas and countries that people are willing to pay for typhoid vaccine [25
], schools with past experience of charging a fee for vaccines did not believe a school-based vaccination campaign was very promising. In order to address concerns of parents and school administration, it is therefore worthwhile to focus on increasing awareness in the population about the disease, its consequences, and importance of a vaccine. A well-designed social mobilization campaign targeting decision makers would also help increase acceptance of vaccines in schools.
There are several limitations in this school census. First of all, there were refusals from 58 schools, which may have biased the data as they represent 5% of all schools surveyed. Secondly, the school census could not obtain some important information. There were no records or information from schools regarding vaccination coverage for vaccination campaigns carried out at the schools. The procurement system for vaccines with a user fee was also not elucidated during the interviews. Provided that there is a user fee, vaccines may be priced at private market cost or may have little subsidy. As vaccine cost is a burden to many parents, provision of vaccines at subsidized price through schools will increase coverage. Third, as we interviewed the school administrators regarding the parents' willingness to participate, information may not truly reflect parents' point of views. Further, our respondents for interview varied from owner of the school to a teacher incharge. Therefore the responses may have varied by the characteristics of the respondent and their individual association with the establishment. Lastly, this census was conducted to identify the baseline for the pilot school-based typhoid vaccination program and there were no pre-determined scientific consideration for analysis.