More media circuses about vaccinating children with MMR, and the "association" of MMR vaccination with autism and inflammatory bowel disease. Frightening stuff, but much heat and little light. One of Bandolier 's informal team asked us to look out the evidence. There's much, mostly from exemplary public health work in Finland, and it shows conclusively that vaccination with MMR does not cause autism or bowel problems.
In Finland, much of the push for immunisation came from the large number of recruits or conscript soldiers who fell ill. In the days before immunisation more than a quarter of them had clinical mumps, and a third of them had orchitis, a quarter of those had bilateral orchitis and a quarter of them were rendered sterile. Mumps contracted during military service was a major cause of infertility, and a major cause of hearing impairment in children, as well as later deafness.
Rubella was also common, with about 1 case per 1000 people a year. Again there was an association with hearing impairment in children, and congenital rubella syndrome was a problem. As a reminder to those of us who forget that infectious diseases are merely inconvenient, when rubella infection occurs during pregnancy, especially during the first trimester, foetal infection is likely and often causes congenital rubella syndrome (CRS), resulting in abortions, miscarriages, stillbirths, and severe birth defects. Up to 20% of the infants born to mothers infected during the first half of pregnancy have CRS. The most common congenital defects are cataracts, heart disease, sensorineural deafness, and mental retardation.
Programme and follow up
In 1982 a major effort was put into vaccinations using a triple MMR vaccine (almost always that produced by Merck "&" Co, which is known by different names in different parts of the world). The programme involves 1000 child health centres, catch up programmes, and military recruits. Two million people (40% of the population) have received 3.5 million doses, and coverage is over 95%.
A national reporting system for mumps, measles and rubella is in force, policed by the National Public Health Institute. Serological confirmation of reported cases has been a requirement since 1987.
The number of cases of measles ran at 15,000-15,000 in the decades before the introduction of a vaccination programme. Since 1985 the number of cases was tiny, and in 1996 fell to zero . For mumps and rubella, the number of cases was a few thousands to a few tens of thousands a year, but fell sharply after the introduction of MMR to a few hundreds or tens (Figure 1), and since 1997 there have been no cases except cases imported from outside Finland .
Harm from vaccines
Measuring the elimination of disease by using vaccination programmes is one epidemiological problem. A harder one, and where the heat is generated, comes from the epidemiological challenge of showing that the vaccines themselves cause no serious harm. Common minor and reversible harm (fever, malaise) is one facet of the problem. Rare major and irreversible harm (autism, inflammatory bowel disease) is another.
In order to tackle both in an effective way, the Finnish National Board of Health and Public Health started a long-term country-wide surveillance system to detect serious adverse events associated with MMR . The system was designed to identify all events that were too rare to have been identified in a blind crossover study on 1200 twins. A series of seminars was held for public health and other involved professionals before the project started, and materials were made available in both official languages (Finnish and Swedish). There was also public information.
In the event of a possible serious adverse event (defined as any temporal association without limit of time between MMR and a life-threatening disorder, triggering of a chronic disease or hospital admission), then a form was completed forwarded to a central office with a serum sample. A second form and second sample followed two weeks later. Forms, envelopes and collection tubes were available at child health centres and hospitals.
The total number of reported vaccine-associated events in 1.8 million people having 3 million vaccinations was 437. Of these, potentially serious adverse events occurred in 169 people, 79 of whom went to hospital. These 169 people were subject to serious scrutiny.
The details of the potentially serious adverse events are shown in the Table. About half the reported adverse events could be ascribed to other factors (like other vaccinations given with MMR) on clinical, serological and epidemiological analyses. No event had an incidence of more than 1 case per 100,000 doses of vaccine.
There were no cases of autism, and no cases of ulcerative colitis, Crohn's disease or any chronic disorder affecting the gastrointestinal tract.
Table: MMR vaccination in Finland - all potentially serious adverse events in 1.8 ,illion people given 3 million doese of vaccine.
Other evidence rejecting a causal link between MMR vaccine and autism comes from a retrospective study of about 500 cases of autism in North Thames . Records of children with autistic disorders born since 1979were identified, and their clinical details examined. Immunisation data was obtained from a separate computerised register.
The results showed that the incidence of autism (core autism, atypical autism and Asperger's syndrome) rose from low levels in 1979 to much higher levels by the early 1990s. This was a continuous trend, with no evidence of a step-up after the introduction of MMR in 1987. There were no differences in age at diagnosis between those vaccinated with MMR and those not vaccinated. No other analysis showed any association between autism and MMR vaccination. There was no temporal clustering between MMR vaccination and diagnosis of autism.
Whenever we take a child to be vaccinated, we are aware that there is a balance between benefit and harm, just like any medical intervention. The potential for harm from infectious disease like measles, mumps and rubella is significant. Unvaccinated children contracting these diseases have a potential for serious and long-lasting harm. The history of mumps and rubella in Finland  makes chilling reading as the litany of harm from these infectious diseases unfolded.
The effects of measles can be seen from a recent outbreak in Ireland. The Irish National Disease Surveillance Centre reports more than 1220 cases of measles in 2000; two children in north Dublin have died. There is also a report of an outbreak in the Netherlands. More than 2300 cases arose in a community philosophically opposed to vaccination. Three children have, 53 hospitalised, four developed encephalitis. That's one death for every 700 children infected with measles in countries with some of the best healthcare in the world.
By contrast, harm is much, much less common. In the Finnish study  there was no more than three adverse effects for every 100,000 doses of vaccines. While even these adverse effects were best avoided, the balance between benefit and harm is very much on the side of benefit. There were no cases of autism in 1.8 million people immunised. There were no cases of inflammatory bowel disease. The rule of three ( Bandolier 23 ) tells us that we can be 95% confident that autism or inflammatory bowel disease occur no more frequently than 1 in 600,000 cases of MMR vaccination.
And this is good, solid evidence. The study was prospective. The study was comprehensive. The study had a long (14 year) follow up. The study had no artificial cut point for parent or professional to link any childhood problem with MMR vaccination. And it was big, and supported by other epidemiological studies .
The evidence taken to link MMR to autism and inflammatory bowel disease  concerned 12 children referred because of loss of skills plus gastrointestinal problems. This was associated with MMR by the parents in eight cases. The authors themselves say that they did not prove a link, which would be difficult, since anyway eight or nine out of every 10 children is vaccinated with MMR.
Finland is a small enough county that it is possible to be in touch with heroes (sharing Sibelius's drinking haunts, or sniffing the tobacco in Mannerheim's study). Perhaps the lesson is that we should take people's obvious concerns about adverse effects of healthcare seriously. People would respect healthcare more if larger prospective studies were done to collect rare but serious harm information. Bear in mind also that there is more than one MMR vaccine. The Finnish study relates to only one of them. One vaccine was withdrawn in the early 1990s , and the UK epidemiological study  does not give the make of vaccines used.
1 H Peltola et al. No measles in Finland. Lancet 1997 350: 1364-5.
2 H Peltola et al. Mumps and rubella eliminated from Finland. JAMA 2000 284: 2643-2647.
3 A Patja et al. Serious adverse events after measles-mumps-rubella vaccination during a fourteen-year prospective follow up. Pediatric Infectious Diseases Journal 2000 19: 1127-1134.
4 B Taylor et al. Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a causal association. Lancet 1999 353: 2026-2029.
5 AJ Wakefield et al. Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet 351: 637-641.
6 MR Kiln. Autism, inflammatory bowel disease, and MMR vaccine. Lancet 1998 351:1358.