Daniel Goldman

Archives for February 2018

The Epidemiology of Gun Related Violence

By Leave a Comment

Second Amendment Banner Art

Epidemiology of gun related violence might seem odd, but violent deaths, including those associated with guns, should be considered a disease. We should therefore use a similar methodology used in treating disease, to identify potential interventions for gun related homicides. Gun control laws are one proposed intervention, or class of interventions. Like with all proposed interventions, we need to test for efficacy and safety. Efficacy requires defining an outcome, and that outcome must be reasonable. We must also take into account what is known as all-cause mortality.

All-cause mortality is an interesting concept and an important one. Consider the case of the health consequences of smoking. We could look at the cancer risk associated with smoking, the relationship between COPD or heart attacks and smoking, and so on. Or we can look at the total estimated impact of smoking on the risk of death. One advantage that this approach has is that it can capture causes of death associated with the exposure (or intervention in the case where you’re testing the efficacy and safety of a new treatment), even ones that may not yet have been considered.

Similarly, when we’re considering whether or not to implement a policy like gun control, which is targeted at reducing homicide rates, we need to consider all-cause mortality here as well. By doing so, we can capture the effects of substitutes for guns in violent deaths, police deaths and shootings, which may result from an increase in people trying to purchase guns on the black market, and even other effects of gun control laws on mortality, whether we have thought of what they might be or not.

This practice may seem like common sense, as it is used so often in health sciences, but a quick look at a major systematic review published in 2016 shows that all-cause mortality is being ignored. In fact, the study actively removed any study which looked at all-cause mortality (Systematic Review on Gun Control). The study did indicate that there seemed to be some evidence that there was a reduction in gun related deaths caused by increased gun control, but ignored what other potential impacts it could have. That is not acceptable. Any research conducted on gun control must include all-cause mortality. We must treat gun related deaths and proposed interventions in the same way that we would study any public health concern.

Further Reading

 

The post The Epidemiology of Gun Related Violence appeared first on Politicoid.

On the Scientific Perception and Reality of Whooping Cough

By Leave a Comment

The following is a draft research proposal to investigate the scientific perception and reality of the efficacy and efficiency of B. pertussis vaccines and the prevalence of asymptomatic infections. I have written extensively, both on The Spiritual Anthropologist, and over at Politicoid, on the topics of vaccine efficacy and science, especially concerning issues with the way vaccine efficacy is researched and on the potential threat of asymptomatic infections, and have been striving to publish new research, rather than just summarizing existing research.

Systematic Review

Although vaccines against B. pertussis have been around for decades, it has only been in the last decade or so that concerns about the ability for the vaccine to prevent infection have arisen. A small clinical trial involving human primates suggests that the acellular B. pertussis vaccine prevents against clinical symptoms, but may not protect against colonization and transmission of the pathogen. Furthermore, in an observational study, B. pertussis infection was confirmed in Israeli children vaccinated with the whole cell B. pertussis vaccine. Both of these findings bring into question whether or not there is a significant ability for these vaccines to prevent colonization and transmission, or if they can only prevent clinical symptoms. Why did it take so long for the medical community to identify weaknesses in these vaccines? Phase I and II clinical trials should have been able to identify how well the vaccine could prevent the spread of infection. Did B. pertussis evolve to become less affected by the vaccines? Did the researchers fail to consider whether or not the vaccines could prevent infection? Or were the methods that they used incorrect?

One indication that the final situation is actually involved is that at least some studies, which claim that the vaccines are effective at preventing infection, actually can only measure the ability to prevent clinical symptoms. These studies look at case reports or gather their data from patients who have gone to seek medical attention for symptoms. An asymptomatic infection is unlikely to result in a visit to a doctor or hospital, and therefore studies which choose their sample from patients seeking medical attention, rather than from the community as a whole, or which look at symptom prevalence, rather than identifying infection using culture or PCR tests, will only be able to identify how well the vaccine does its job of preventing disease.

A systematic review of available studies on B. pertussis vaccines can help answer the question of whether or not study methodology is at fault. The proposed systematic review will start by searching for all studies which investigate the efficacy and efficiency of B. pertussis vaccines, either whole cell or acellular, whether currently in use or not. Duplicates will be removed, and the studies will be sorted into categories. The categories will be as follows: investigates efficacy, investigates efficiency, sample drawn from general population, sample drawn from those seeking medical attention, identifies symptoms only, identifies infection, comments on ability to prevent infection, and comments on ability to prevent disease. In this way, the study can investigate what fraction of studies have conflated ability to prevent disease with the ability to prevent colonization and transmission. While it is not crucial to the question at hand, a quality analysis will also be conducted to see if there is a relationship between study quality and tendency to conflate these concepts.

The full systematic review will involve two readers. I will not be involved in the review process as the amount of material I have written and read on this specific topic may skew my analysis of the papers.

Model Construction & Parameter Estimation

The basic SVIR model is a useful tool for estimating the dynamics of how infections spread, under various vaccination regimes. However, the model does not take into account how asymptomatic infections, either due to natural partial immunity, or due to partial immunity derived from vaccination, affects epidemics. Parameters that will need to be studied or assumed, before the model becomes viable, include how likely an unvaccinated person is to become asymptomatic upon infection, how likely a vaccinated person is to become asymptomatic upon infection, and how symptoms affect contact rates between healthy and unhealthy individuals.

A key assumption of the model that will be constructed is that appearance of symptoms influences how likely a healthy and infected person are to come into contact with one another. This assumption is based on theories about disease avoidance. Kouznetsova et al. 2011 indicate that such behavior is not limited to actual contagious diseases, but that people will avoid anyone who even appears to be contagious [1]Kouznetsova, Daria, Richard J. Stevenson, Megan J. Oaten, and Trevor I. Case. “Disease-avoidant behaviour and its consequences.” Psychology & Health 27, no. 4 (2012): 491-506. doi:10.1080/08870446.2011.603424.. Risk aversion also occurs in sick individuals. Even though many Americans go to work sick, a significant fraction do stay home or take additional precautionary measures. According to a survey conducted by nsf.org, at least 26% of workers admit to going to work sick. 17% of women and 33% of men admit to always going to work sick. However, those numbers still indicate that a significant number of individuals do stay home, especially if a doctor tells the person to do so.

These results indicate that the rate of contact between healthy and sick individuals is a function of how obvious it is that the person is sick and potentially contagious. Asymptomatic carriers and those with minor symptoms do not appear to be contagious, and might not know that they are sick, and therefore there may be little to no risk avoidance behavior. Ideally, this concept would be modeled using a system of partial differential equations, but ordinary differential equations are much easier to manipulate.

Observational Study

Once parameter estimates are made and the model constructed and evaluated, it will be time to obtain empirical justification of the model. If the model is a reasonable one, it should be able to give us an estimate of the current rate of B. pertussis infection. Even if the model itself fails, the results from obtaining actual estimates for the incidence rates of infection, within the general population, and certain at risk sub-populations, will be useful, both for adjusting the model and for updating public health policy.

The basis of the observational study is a similar study conducted in China. The cross-sectional study attempted to identify the prevalence of asymptomatic infections among school aged students (aged 7 – 15) in four different counties in China. Both culture and PCR testing was used, and sampled were collected between July and September of 2011.

Bibliography

  1. Kouznetsova, Daria, Richard J. Stevenson, Megan J. Oaten, and Trevor I. Case. “Disease-avoidant behaviour and its consequences.” Psychology & Health27, no. 4 (2012): 491-506. doi:10.1080/08870446.2011.603424.
  2. Kurzban, Robert, and Mark R. Leary. “Evolutionary origins of stigmatization: The functions of social exclusion.” Psychological Bulletin 127, no. 2 (2001): 187-208. doi:10.1037/0033-2909.127.2.187.
  3. Srugo, Isaac. “Pertussis Infection in Fully Vaccinated Children in Day-Care Centers, Israel.” Emerging Infectious Diseases6, no. 5 (2000): 526-29. doi:10.3201/eid0605.000512.
  4. Fulton, T. Roice, Varun K. Phadke, Walter A. Orenstein, Alan R. Hinman, Wayne D. Johnson, and Saad B. Omer. “Protective Effect of Contemporary Pertussis Vaccines: A Systematic Review and Meta-analysis.” Clinical Infectious Diseases 62, no. 9 (2016): 1100-110. doi:10.1093/cid/ciw051.
  5. Warfel, J. M., L. I. Zimmerman, and T. J. Merkel. “Acellular pertussis vaccines protect against disease but fail to prevent infection and transmission in a nonhuman primate model.” Proceedings of the National Academy of Sciences 111, no. 2 (2013): 787-92. doi:10.1073/pnas.1314688110.
  6. Zhang, Qi, Zundong Yin, Yixing Li, Huiming Luo, Zhujun Shao, Yuan Gao, Li Xu, Biao Kan, Shan Lu, Yan Zhang, Manshi Li, Meizhen Liu, Pingping Yao, Zhixian Zhao, and Qiushui He. “Prevalence of asymptomatic Bordetella pertussis and Bordetella parapertussis infections among school children in China as determined by pooled real-time PCR: A cross-sectional study.” Scandinavian Journal of Infectious Diseases 46, no. 4 (2014): 280-87. doi:10.3109/00365548.2013.878034.

Webpages

https://www.nsf.org/newsroom_pdf/Flu_in_the_workplace_(final).pdf

http://www.nsf.org/consumer-resources/studies-surveys-infographics/consumer-survey-results/flu-in-workplace

 

References   [ + ]

1. ↑ Kouznetsova, Daria, Richard J. Stevenson, Megan J. Oaten, and Trevor I. Case. “Disease-avoidant behaviour and its consequences.” Psychology & Health 27, no. 4 (2012): 491-506. doi:10.1080/08870446.2011.603424.

The post On the Scientific Perception and Reality of Whooping Cough appeared first on The Spiritual Anthropologist.