Taking stock: Shabir Madhi says it's essential to know how well the immunisation programme is working.
Taking stock: Shabir Madhi says it's essential to know how well the immunisation programme is working.

South Africa has been praised for introducing new and expensive vaccines into government’s expanded programme on immunisation (EPI), which provides free vaccines in the public sector. But experts warn that the full impact of this can’t be known because the data about how many citizens receive these vaccines is flawed.  

In 2009 South Africa was the first country on the continent to introduced two new vaccines – Pevnar, which protects against pneumonia and meningitis caused by pneumococcus bacteria, and Rotarix, which protects against diarrheal disease caused by the rotavirus. Prevnar costs the government about R 2000 per child according to the head of the National Institutes of Communicable Diseases, Shabir Madhi.  

Last year, South Africa made the new human papilloma virus vaccine Cervarix available to all school girls aged between nine and fourteen. The only other African country to have done this was Zambia, according to the World Health Organisation (WHO). 

But coverage estimates differ widely between sources and all figures have been contested. For example the department of health estimates that 82.8% of children received the measles vaccine in 2014 but the WHO says it was 60%. 
 

In 2012, the WHO claimed 64% of South African children were fully immunised (in other words they received all seven vaccines in the EPI) but the health department estimated the figure to be over 30% higher (96%).
 “The truth is we actually don’t know how many children are immunised,” says Madhi. 

No Survey
No nation-wide vaccine coverage survey has been done since 1998, he says, because it is very expensive and the current methods of acquiring this data aren’t an accurate substitute. Madhi says that the reason for the delay is that a survey would cost the government at least R25-million.  

“But, if you look at it in context, we spend about R1.2-billion just on procuring vaccines each year. If you take into account administrative and staffing costs, the figure will probably double. For us to spend that amount of money without really understanding whether we are actually getting the benefit is problematic.”

The data issued by the government and international sources is “just speculative”, Madhi says.
The government currently bases its estimates on data gathered from the district health management information system (DHIS), which was began in 2000 and covers hundreds of different indicators, including HIV and lifestyle diseases information, routinely submitted by districts. All immunisations logged at facilities are divided by the estimated child population in that district.  

“The problem with this is that immunisation information probably accounts for less than one percent of this survey and it lacks any specificity.”

The immunisation cards used at public facilities were changed only two years ago to reflect the new vaccines added to the EPI in 2009. 

Accuracy
Madhi says staff in clinics have difficulty recording this information accurately, especially if children are still using the previous card.  

This data is also aimed at providing information on a national level and isn’t structured to assess the sub-district level. “And in a country like South Africa, where there is inequity in terms of accessing even primary healthcare, it is pivotal we know details about exactly where people aren’t being immunised so we can find out why.” 

Another problem is that the DHIS relies on Statistics South Africa data for its denominators, or estimates about population size, which fluctuate in time and are problematic themselves. 

Because of inaccurate population estimates some of the DHIS vaccine coverage data is overestimated.
For instance all five Gauteng districts have coverage rates of more than 100%, “which is obviously impossible”. 

According to Helen Rees, who is part of the team which advises the health minister on immunisation, these estimates can be partly attributed to the fact that there is “huge migration from rural to urban areas as well as into the country from elsewhere in Africa”.  

Madhi and the health department’s deputy director general for maternal health and HIV, Yogan Pillay, say, the WHO statistics are also questionable. 

According to Pillay, WHO calculates its estimates using information from 1998 because it does not accept the government’s data as being accurate.
“But using this data is erroneous because much has changed in the fifteen years since that survey,” he says.

The 2009 measles outbreak where there were 20 000 confirmed cases, and possibly three times more that were undiagnosed, shows there is a need to improve coverage urgently, he says. “But we can’t answer the question about where we need to increase resource allocation to get the improvement.” 

He warns that efforts to encourage citizens to vaccinate their children, are also affected because “we aren’t sure where to focus our efforts”.

He says vaccine shortages at public health facilities and other barriers to immunisation experienced by patients can be addressed only by a detailed survey.
“We suspect that shortages happen more in rural settings, but a survey will allow us to really understand the magnitude of the problem and concentrate our effort on getting vaccines to the people who need them most.” 

It’s especially important to have high immunisation coverage in low-resourced and rural settings because “this is where people usually have the least access to curative healthcare and this is where vaccines can really save many lives”.

Pillay says a survey is planned for 2016 if the funds become available. “But we need to find the money first. It will cost at least R20-million and it is unclear at this stage whether we will get the budget from treasury or if a donor will step in to help.”

The cost is nominal, says Madhi, when “you consider that next to clean water, vaccination is the most cost effective strategy in existence to improve health, especially in children”.