Graphical proof that vaccines work (with sources)
The graphs in this article (showing disease incidence over time relating to several diseases and several countries) make a compelling case for vaccine effectiveness.
Proponents of the anti-vaccine movement will often plot death rates over time as “proof” that vaccines were introduced at a time when they had a negligible effect. However, mortality rates are the incorrect metric when evaluating vaccine effectiveness. Here’s why:
- Disease incidence is the correct metric. Since vaccines are designed to prevent the incidence of disease, rates of disease (morbidity) are the correct metric for evaluating vaccine effectiveness. If vaccines were a drug to treat disease, then case-fatality rates would be the correct metric for evaluation. Measuring vaccine effectiveness by looking at mortality rates only is a bit like measuring contraceptive effectiveness by measuring abortion rates only — it only tells a tiny part of the story.
- Mortality rates do not reflect the true burden of the disease. Survivors of vaccine preventable disease can be maimed for life — think polio and paralysis, mumps and deafness, measles and mental disability, rubella and birth defects. Looking at mortality rates only misses the true burden of the disease. For example, before the introduction of the measles vaccine in the US in the 1960s, there were “only” 400–700 measles deaths a year. However, there were 4,000 cases of measles encephalitis a year (many of which led to neurological complications such as blindness, deafness and mental disability), 150,000 cases of respiratory complications, and 48,000 measles-related hospitalizations a year (which placed a huge financial and emotional burden on families) . Together, these figures added to the motivation to eradicate the disease.
- Numerous factors affect mortality rates. Mortality rates are influenced by socioeconomic status, access to health care, quality of treatment and a wide range of other factors. As these factors change, mortality rates inevitably change. By contrast, as you can see from the graphs below, vaccine preventable disease incidence remained largely constant until the introduction of related vaccines. This may be because airborne diseases (like measles, mumps and rubella) can infect anyone, regardless of socioeconomic status.
- Mortality rates are not more accurate than morbidity rates. Those opposed to vaccination often use the “disease diagnoses are inaccurate” argument as a reason to look at mortality data only. However, the mechanism by which a doctor’s diagnosis suddenly becomes accurate when a patient dies, when he or she was not competent at diagnosing the disease in the first place, is not explained. In fact, death certificates will often be based on preceding diagnoses. Furthermore, as other researchers have noted, disease cases are more likely to be accurate as death certificates were only standardized from the 1960s onward .
All of the graphs below are custom compiled for this article based on data made available via the United States Census Bureau, CDC and Public Health of England websites. I encourage you to double check the sources (included as links below) and compile your own graphs in Excel to double check these trends.
You are welcome to share and re-blog these graphs in any way you like in your fight against vaccine misinformation. The data has been double-checked against sources, however if you do notice any inconsistencies or have additional data to share, please contact me via Twitter.
In the pre-vaccination era, more than 90% of Americans were infected with measles by the age of 15 (roughly 4 million people per year). Not all cases were reported to the public health system and reported cases averaged 542,000 per year from 1956 to 1960 .
The following graph shows population-adjusted reported measles cases provided by the United States Census Bureau 1912–2001 Reportable Disease Report. Unfortunately, data only goes as far back as 1912. As you can see below, measles used to peak in 2–3 year epidemics until the introduction of measles vaccines in the 1960s.
The CDC also has data on the total number of reported cases from 1950–2011. Unlike the data from the Census Bureau, this data is not population-adjusted but reflects the total number of reported cases over that time period.
More US graphs on measles
In England and Wales, the measles vaccine was introduced in 1968. Initial vaccination coverage was low, but it slowly increased to a level of approximately 80% in 1988. The introduction of the vaccine had an immediate effect on the incidence of measles, shown below. Measles data from 1940 until 2013 is provided by the National Archives (Public Health England) website, which also provides data on vaccine uptake.
Some studies have noted that measles notifications are not 100% accurate. However, this observation is based on a 1991–1993 pilot study comparing notifications to lab-confirmed cases during that same time period. In the pre-vaccination period, when measles infection was universal, the number of actual cases would have been much higher than the number of reported cases, as only a fraction were reported to the public health care system during that time.
More UK graphs on measles
- Annual measles notifications and vaccine coverage in England and Wales, 1950–2001 from Evolution of Surveillance of Measles, Mumps, and Rubella in England and Wales
- Annual measles notifications and vaccine coverage, England and Wales 1950–2009
- Annual measles notifications & vaccine coverage in England and Wales 1950–2000 (PDF)
Population-adjusted polio data is available via the United States Census Bureau 1912–2001 Reportable Disease Report. When the vaccine was introduced in 1955, cases dropped dramatically and have been close to zero since 1966.
Data on population-adjusted polio death rates in the United States (ie. the number of polio deaths per 100,000 population) is available on the CDC website by going through mortality data on “Vital Statistics” PDFs per year. Here is a look at polio death rates from 1910–1975 (unfortunately, no pre-1910 polio data is available). The polio epidemic in 1916 had the highest death rate recorded (10.5 per 100,000 population) and matches the epidemic lines in figure 2.1 above.
If you compare polio cases and deaths in the 1916 and 1952 polio epidemics (figures 2.1 and 2.2 above), case-fatality rates were dropping. This is not surprising as medical care would have improved over a span of 37 years. However, despite falling case-fatality rates, population-adjusted polio cases were steadily rising from 1943 until the introduction of the vaccine in 1955, so that death rates were also on the increase over that time period. Here is a closer look at polio death rates from 1917 until 1975 which illustrates this phenomenon:
More US graphs on polio
Data on paralytic polio, non-paralytic polio and polio deaths from 1912 in England and Wales is available on the National Archives (Public Health England) website. Cases and deaths declined dramatically since the introduction of the polio vaccine in 1956.
More UK graphs on polio
- Polio notifications in England and Wales (1912–2006) (PDF, page 2)
Population-adjusted diphtheria data is available via the United States Census Bureau 1912–2001 Reportable Disease Report. The diphtheria vaccine was introduced in 1923 and caused a rapid decline in cases and deaths. The United States has been diphtheria free since 1977.
The pertussis vaccine was developed in the 1930s and used widely in clinical practice by the mid-1940s. The following graph shows population-adjusted cases, courtesy of the United States Census Bureau 1912–2001 Reportable Disease Report (unfortunately no pre-1922 data is available on pertussis). Current rates of pertussis cases are very low compared to the pre-vaccination era.
The pertussis vaccine was introduced in England and Wales in the 1950s, though the exact date is unclear. Due to vaccine controversies, coverage dropped to around 30% in 1975 which contributed to two major epidemics, shown below. Since the mid 1990s, vaccine coverage has increased to 90% and cases have declined again. Figures on pertussis cases and deaths is available on the National Archives (Public Health England) website.
More UK graphs on pertussis
Haemophilus influenzae type b (Hib)
Before the introduction of Hib immunisation, the estimated annual incidence of invasive Hib disease was 34 per 100,000 children under the age of 5. The Hib conjugate vaccine was introduced in 1992 in the childhood immunisation schedule. Since then, disease incidence has fallen by over 95% in infants under one year of age (the highest risk age group for disease). Data on laboratory confirmed cases from 1990 onward is available on the National Archives (Public Health England) website.
More UK graphs in hib
Hepatitis B, especially if contracted during childhood, can lead to liver cancer, liver failure and death. A hepatitis B vaccine became available in the US in 1982 for high risk cases, but had limited uptake. Since the vaccine was added to the infant schedule in 1991, cases declined dramatically. The following graph, showing reported cases from 1996 until 2011, is based on data made available by the CDC.
Data on hepatitis B deaths is available from 1979 onward. People who die from hepatitis B typically do so after chronic infection. Thus, it makes sense that a decline in hepatitis B deaths is evident several years after the introduction of the vaccine.
Chickenpox only became a reportable disease in the United States in 1972. Chickenpox can cause complications such as pneumonia and encephalitis. Furthermore, people who contract chickenpox have a 30% chance of developing shingles later on in life. The following data shows reported chickenpox cases from 1972 until 2011, based on data provided by the CDC:
The following graph shows reported chickenpox deaths in the United States from 1972 until 2007, also provided by the CDC.
Others vaccine preventable diseases
I hope to expand the list of graphs on this page but have not been able to find historical data on all vaccine preventable diseases, specifically:
- United States: Tetanus (no pre-1950 morbidity data), Mumps (no pre-1968 data), Rubella (no pre-1966 data), Hib (no pre-1991 data)
- United Kingdom: Diphtheria (no pre-1986 notifications), Hepatitis B (no pre-1990 data), Mumps (no pre-1989 data), Rubella and CRS (no pre-1989 data), Tetanus (no pre-1969 data), Chickenpox (no data)
If you come across historical data for additional diseases or additional countries, please contact me and I will gladly add to this page.