House on Fire (23 page)

We tracked other metrics besides outbreaks. We looked at numbers of deaths and numbers of cases for some information, putting them into rates per hundred thousand people or per month, week, or even day. Sometimes, to integrate data from search weeks with data from other weeks, we would use moving averages, or the average number of new outbreaks or cases per week by including two weeks before and two weeks after the week being examined. All of these were attempts to see
the problem from different perspectives, and sometimes the numbers didn't easily compare. What were the best measures? If I wanted to have a guess at the size of the problem in order to plan for work assignments, I would look at pending outbreaks. If I wanted to know whether we were gaining on the problem, I would compare new outbreaks and contained outbreaks for a geographic area. But if I wanted to see the full fury of what we were up against, nothing was as telling as new cases per day. That brought the problem down to the people and families hurt forever on that day. At the height of the conflagration, there were fifteen hundred new cases of smallpox a day in Bihar—one new case every minute.

While the hope had been to conduct a search every month, the scheduling didn't quite work out that way. The fourth search was conducted February 4 to 9, 1974, and the fifth search from March 18 to 23. The sixth search spanned the last few days of April into the first days of May. Through the twenty months of the intensified campaign, there were sometimes only three weeks between searches; once, there was a seven-week gap. Therefore, some of the numbers that follow will refer to specific searches while at times, for clarity, the reports will indicate the average number of outbreaks found or contained by week during a given period.

In Uttar Pradesh, the number of new outbreaks found in each search was between 330 and 750, and the number of pending outbreaks continued to increase, so that 1,759 pending outbreaks were on the rolls in April 1974. Previous experience had given us no way of comprehending the work involved in dealing with over 1,700 pending outbreaks in a single state. Yet the situation was even worse in Bihar state.

The third search, in December 1973, had revealed slightly over 2,600 new cases of smallpox in Bihar. During the fourth search, in February 1974, 5,821 new cases, involving all districts, were reported. In terms of new villages with smallpox, the number had increased by 1,170. (Large villages could have more than one outbreak, but for practical purposes, the number of new villages infected was nearly the same as the number of new outbreaks.)

By March the situation had worsened. The number of new outbreaks reported that month averaged more than 600 per week, increasing to 1,000 some weeks in Bihar alone. This was almost twice the number of outbreaks being contained, so the number of pending outbreaks continued to increase. During the week of the fifth search, the discovery of 2,345 new outbreaks in Bihar increased the total pending outbreaks to 3,683 after the contained outbreaks were removed from the list.
Figure 12
shows the number of new and contained outbreaks in Bihar during the first four months of 1974, combining information from the monthly searches, secondary and tertiary surveillance systems, and the traditional passive reporting system, which continued to operate throughout the smallpox eradication program.

Figure 12.
Average number of new and contained outbreaks per week, Bihar, India, January to April 1974

The numbers of cases and outbreaks climbed partly because surveillance kept improving. The monthly searches involved a frenzy of activity in districts and PHCs as health staff were diverted to the task of finding smallpox. Of course the remaining health workers were then left with increased workloads. Rarely are these workers acknowledged in accounts of the program, but they were absolutely key and should be counted as part of the vast army that comprised the smallpox warriors.

Every month saw innovations in search techniques. Early in the program the searchers began supplementing their earlier protocol by spending more time questioning children. Then the surveillance system
moved from village searches—in which searchers would talk to key informants such as teachers, students, and village leaders—to going house to house in selected parts of the village. Eventually, searches developed to the point of going to every house.

The proficiency of the searchers grew month by month, both as a group and individually. Some searchers were more diligent than others, and switching searchers' assignments frequently resulted in detection of cases that had not been reported the previous month. That of course meant that those outbreaks had been free to expand.

The workers' objectives changed through the months, too. Early on, their objective might have been to correctly complete a form, but then they began to take pride in finding cases that they might have missed months earlier. Everyone was getting caught up in the challenge, and the concept of a team approach was infectious.

Surveillance also improved markedly as the program developed. At first, the surveillance approach was an extension of approaches used in Africa. Gradually, however, smallpox workers discovered how to do better surveillance, including how to identify both the population groups most likely to have smallpox and those most effective in hiding smallpox cases. The discoveries of individual workers, reported in monthly meetings, became the basis for improving surveillance overall. Secondary and then tertiary search techniques evolved, based on what was being missed by the six-day searches. The techniques were intended to be redundant, thereby increasing the likelihood of recognizing the presence of the virus. During the weeks between main searches, supplementary searches became standard in areas of high suspicion. There was a sense—later confirmed by the data—that the rate of identifying outbreaks and existing cases was increasing.

With experience it became possible to draw up guidelines for supplementary surveillance. Secondary search techniques involved sending teams to wherever crowds assembled—markets, fairs, religious ceremonies. Smallpox workers would go through the crowds with recognition cards, asking if anyone had seen a person with smallpox. The teams submitted positive reports to their supervisors, including the name of the person's village and whatever other information they could garner.

Tertiary searches were started for groups that were not integrated into the local community and thus not likely to be found in village searches. These included the Harijan communities (the “untouchables”) and occupation groups such as road builders and kiln workers, who moved from place to place in search of work. Beggars were usually found at markets, in urban areas, or at railway stations; less often in villages.

The high degree of mobility in India presented another difficulty. Millions of people are on the trains at any one time, not to mention buses, trucks, and other vehicles. In addition to the usual reasons for travel, such as commerce and family visits, people would travel to fairs and religious meetings, which could draw more than twenty-five thousand visitors to a single site. The previous October, in Bihar alone, almost 8 million people had been expected to visit fairs. A single religious
mela
(festival) in Uttar Pradesh meant over 5 million visitors converging on a small geographic area. Special teams had to be deployed to search and vaccinate in these large congregations.

By January 1974, the searches that had become so useful in smallpoxendemic states were extended to the entire country. Searches of the nonendemic states were conducted every three months. To visit 100 million houses in six days' time, an army of workers had to be organized. The procedures that had become routine in the smallpox-endemic states now had to be replicated in the other states, districts, and PHCs. Tens of thousands of searchers, supervisors, and evaluators needed to be trained. It constituted one of the largest operations in public health history.

Every three months, the country obtained a snapshot of where the smallpox virus was hiding. When previously unknown cases were found in non-endemic areas, the teams had the capacity to respond quickly to avoid large outbreaks. The presence of cases in non-endemic states meant that the infection had been imported from an endemic state. The quarterly searches of the non-endemic states not only stopped spread from the endemic states but also indicated the origin of the cases, thus helping to pinpoint the problem areas in endemic states. This in turn put more pressure on those districts to protect both themselves and the rest of India.

Improving containment was the key to responding to the soaring numbers. In early 1974, the huge gap between new outbreaks and contained outbreaks was beginning to narrow, but only slightly; yet knowledge of the field situation was improving at PHC, district, and state offices. By this time, the national operating instructions for the smallpox program called for a supervisor to actually visit each outbreak to evaluate the work of the containment team. That didn't happen right away, of course, because the ideal was always ahead of the actual. Also in early 1974, with a high degree of hope, containment teams were directed to return to the practice of vaccinating twenty households around each house with smallpox.

The number of outbreaks was so large that at first the containment teams were unable to respond adequately. They would vaccinate the affected households and immediately surrounding households but would then have to go to the next outbreak. Moreover, as the early months of 1974 progressed, the large number of outbreaks meant that even though there were more containment teams, they might not be able to visit the majority of outbreaks before the next search added even more outbreaks to the list. PHC and district medical officers were forced to prioritize outbreaks. With so many outbreaks, even basic record keeping, a minimum requirement, was a challenge.

Besides the rapid increase in outbreaks, the labor-intensive nature of outbreak containment and the difficulty of continuing to train more teams left the program constantly running behind in containment capacity. For each outbreak, an investigation was required to determine the origin of the outbreak and to notify the PHC that had exported the virus. Vaccinations, which eventually included census operations, had to be organized. The search had to be continued for weeks after the last case. Finally, the containment team had to close out the outbreak once the last patient had recovered. The amount of work, record keeping, and supervision for one outbreak was prodigious. Even after containment work had ceased, repeated visits were required for four to six weeks afterward to be certain no new cases had developed. These visits were
followed by repeated visits from a supervisor to insure that protocols had been properly followed and the outbreak had been terminated.

The concept of a six-foot perimeter of no susceptible people around a smallpox patient proved to be a useful rule of thumb in stopping outbreaks. But, just as a forest fire can jump a firebreak under certain conditions, so the virus could occasionally jump this thin line. This was rare, however. Don Francis reported one of these unusual events. To commemorate the death of a young boy from smallpox, the family invited all of the village children of the same age to their home for a
kheer
(rice pudding) party. As the children were sitting on the front stoop of the house eating their
kheer
, the dead child's grandmother began cleaning the child's sickroom. She brought the bedding out to the front stoop and shook and beat the child's quilt next to where all the children were sitting. Ten days later, these un-immunized children all came down with smallpox.

It was soon discovered that people from outside the immediate family and even outside the neighborhood were expected to visit a person who was sick with smallpox, providing considerable opportunity for the virus to jump the six-foot perimeter. Forbidding visits was simply not enforceable; this rule, interpreted as an offense to the goddess of smallpox, simply drove the practice underground. The solution was to post watch guards at each patient's house with instructions to vaccinate every visitor. The watch guards were provided with journals and were expected to record the names of those they vaccinated, as well as their own food and bathroom breaks. Hiring, training, and deploying watch guards at each house was added to the work of the containment teams.

The numbers involved in just this small aspect of the program once again demonstrate the size of the containment actions. Posting two watch guards per house meant hiring and training ten new people for every outbreak (the average outbreak consisted of five cases). Five thousand pending outbreaks in Bihar meant a new workforce of fifty thousand people. Later in the year, the number of watch guards for each infected house was doubled to four to insure round-the-clock coverage. At the peak of the 1974 smallpox season, eight thousand pending outbreaks alone required between fifty and a hundred thousand watch guards.

Another problem was discovered when it was observed that family members or neighbors of vaccinated people were coming down with smallpox one incubation period later. They had been missed in the vaccination. Many were working in the fields, buying food or petrol, or engaging in other life activities. Some, however, were missed because they intentionally avoided vaccination, mistrusting anyone from the outside, especially government workers. Some were children who had been hidden by their parents.