As part of routine measles case investigations, urine samples and throat swab specimens for virus isolation were collected along with serum samples (0–12 days after rash onset) from patients across Uganda during 2006 through 2009. Infections were confirmed serologically or by virus isolation (4,5). Of the serum samples tested, 1,053 (15%) of 6,999 were positive for measles immunoglobulin (Ig) M; most were collected during 2006 (Figure 1). Twenty-two isolates (37%) were obtained from 59 samples from patients who had IgM against measles virus; 1 isolate was obtained from a patient who did not have IgM against measles virus; and another isolate was obtained from a patient who did not have serologic testing. Virus isolation was successful only in specimens collected within 5 days of rash onset (Table).

All isolates belonged to genotype B3.1 (Figure 2), which had not been previously detected in Uganda (3). Twelve (57%) of 21 sequences obtained were identical (Table) and also identical to isolates from the 2005 measles outbreak in Kenya (Figure 2). However, 9 (43%) of 21 showed neither 100% similarity with the other 12 Ugandan isolates (Table) nor with any other isolate available in GenBank.

Since the inception of Uganda’s 2002–2006 accelerated measles control strategic plan, the number of measles cases in the country has declined dramatically (Figure 1). After the 2003 campaign, virtually no cases of measles occurred in Uganda for 3 years, until the outbreaks in 2006 (1). Our data show that 819 (78%) of 1,053 of the serologically confirmed measles cases for the 4-year surveillance period occurred during the 2006 outbreaks (Figure 1), confirming that the strategic interventions quickly subdued the 2006 transmission cycles. Moreover, the pattern of measles genotypes detected from 2000–2009 (2006–2009 reported in this study) suggests that transmission of the previously endemic genotype D10 in Uganda (3) has been interrupted and replaced by genotype B3.1. However, since molecular surveillance in Uganda only began 10 years ago (3), we cannot ascertain which genotypes were circulating in the country before 2000. By 2001, B3.1 was geographically restricted to the western and central African countries of Ghana, Nigeria, Cameroon, and Sudan (6,7); however, by the end of 2005, it had spread to Kenya, supposedly from Nigeria, where it caused a massive epidemic that was linked to subsequent infections in Europe and the Americas (8).

The fact that most isolates from Uganda were identical to those previously identified in neighboring Kenya indicates that Kenya was the most likely immediate source of the B3.1 viruses presently circulating in Uganda. However, the contribution of other African countries cannot be excluded, because molecular surveillance is still largely lacking in Africa (5).

Only 14 (23%) of 61 patients with laboratory-confirmed measles (whose specimens were screened for measles virus isolates) had been vaccinated, demonstrating a gap in vaccination coverage and possible vaccination failure in some cases and the need for timely catch-up vaccination campaigns even in the low-risk districts. In Uganda, routine childhood vaccination against measles began in 1983, but vaccine coverage was initially too low to interrupt indigenous transmission. The situation was not helped by the rampant poverty, limited health-care infrastructure, internal and external conflicts that have rocked the Great Lakes Region during the past 3 decades, high vaccination drop-out rates, and a high birth rate (second highest in the world, after Niger [9]). These conditions are typical in countries where vaccine-preventable infections remain a major problem (10). Nevertheless, education of Ugandan health workers, politicians, and development partners has been ongoing. This strategy has aroused strong interest and support for disease surveillance and control activities from the political leadership, resulting in the creation of a special budget-line for surveillance (11). Since 2003, routine vaccination has been strengthened by extending primary health care grants to all districts, and the implementation of the “Reaching Every District” strategy (12), among other interventions. Consequently, Uganda has recorded a tremendous rise in vaccination coverage, from 64% in 1997 to >85% by 2007 (1). This aggressive vaccination effort has been instrumental in interrupting indigenous measles strain transmission in the country.

All virus isolates from Uganda during 2000–2002, before accelerated measles control, belonged to genotype D10 (3). At that time, viruses isolated in countries bordering Uganda, such as Kenya, Sudan, and the Democratic Republic of the Congo, belonged to genotypes D4, B3, and B2 (13). However, without enhanced regional surveillance, should genotype D10 be isolated again, it would be difficult to determine whether it had truly been interrupted, without knowing if D10 was presently circulating in other African countries, which could serve as bases from which to reintroduce it into Uganda.

Our results show that, even under difficult circumstances (e.g., poverty), optimal resource allocation and mobilization of political will can interrupt vaccine-preventable diseases in Africa. These data provide molecular evidence that Uganda’s 2002–2006 vaccination strategy was successful in interrupting indigenous measles transmission, but immunity gaps in the population allowed the establishment of an imported virus that was previously confined to western and central Africa. If national immunization programs across the region synchronized their vaccination strategies to eliminate sources of reintroduction, measles could be quickly eliminated from the entire continent. Vaccination success stories have already been noted in several African countries with routine coverage >80% (14,15). Therefore, continued education and cooperation are needed between countries, national policy makers, health care workers, and local communities throughout the continent to win the fight against measles.