Causal Analytics for Applied Risk Analysis Louis Anthony Cox, Jr

Describing quantitatively how large a risk is provides crucial information for helping to set risk management priorities. This chapter applies descriptive analytics to assess the size of the human health risks from a particular source. It continues the theme begun in Chapter 5 of examining human health risks from antibiotic-resistant infectious bacteria, but focuses specifically on long-standing concerns expressed by the public health community, news media, and members of the general public that methicillin-resistant Staphylococcus aureus (MRSA) transmitted from pigs to humans may harm human health. For readers not interested in details of this application, a brief summary is that previous studies of the prevalence and dynamics of swine-associated (ST398) MRSA have sampled MRSA at discrete points in the presumed causal chain leading from swine to human patients, including sampling bacteria from live pigs, retail meats, farm workers, and hospital patients. This chapter integrates available data from several sources to construct a conservative (plausible upper-bound) probability estimate for the quantiative human health harm (MRSA infections and fatalities) arising from ST398-MRSA from pigs. The estimated plausible upper bounds are approximately one excess human infection per year among all U.S. pig farm workers, and one human infection per thirty-one years among the remaining total population of the U.S., assuming that bacteria transmission events not yet observed are possible. The true risks may be smaller (possibly zero for members of the general population). Putting rough numerical bounds on the size of the risk such as “less than one case per 300 million people per decade” can help to engage System 2 thinking (cognitive, slow, deliberative; see Chapter 12) about costs and benefits of risk-reducing interventions; whereas leaving the risk unquantified is more likely to engage System 1 thinking (intuitive, quick, emotional; see Chapter 12), which typically reacts very strongly and adversely to the qualitative idea of contamination of food with invisible, harmful microbes.

Introduction: How Large is the Human Health Risk from MRSA in Swine and Pork?

Recent news stories, and many web sites and blogs, have warned that antibiotic-resistant bacteria from pig farms may threaten public and worker health. One widely discussed concern is that animal antibiotics used on farms create selection pressures that favor the spread of antibiotic-resistant “superbugs,” such as methicillin-resistant Staphylococcus aureus (MRSA) (CBS, 2010). However, these stories have lacked a quantitative discussion of how many excess deaths, treatment failures, or days of illness each year are caused in the United States by MRSA arising from pig (or other livestock) production. Most human cases of MRSA arise from antibiotic use (and inadequate hand washing and infection control) in hospitals, and mainly affect hospitalized patients with severely compromised immune systems. How many cases of MRSA infections and fatalities per year arise among hospital patients, butchers, slaughterhouse workers, farmers, or the general public from livestock operations, meat handling, and consumption therefore remains to be addressed. Responsible risk management is supported best by understanding how large the human health risks are now, and how much they would be changed by proposed interventions.

Potential Human MRSA Hazards Related to Pigs and Pork

This section takes inventory of what is known about adverse human health outcomes which might be caused by exposure to MRSA from swine, and pork and identifies data on the numbers and types of adverse health consequences per year which might be caused by swine-associated (especially, sequence type 398 or ST398) MRSA. The purpose is to identify sources of risk (i.e., hazards) which pose non-negligible human health risks, so that these can be quantified further.

Consumer Exposure to MRSA via Pork Meat Poses Little Risk

Consumption of pork meat raises the logical possibility of three types of exposure and harm from MRSA: invasive infections by MRSA acquired from meat; non-infectious food poisoning by MRSA enterotoxins (as with other strains of S. Aureus); and colonization (with the possibility of subsequent opportunistic infection) by MRSA from pork products. Each of these is briefly considered next. We find that only the last, colonization, appears to pose a potentially significant risk in the U.S.

• 
  Invasive infectionby MRSA consumed in pork products. No case of a livestock-associated (ST398) MRSA infection transmitted via the food supply (or otherwise) has ever been found in the United States (Limbago, 2010). Worldwide, only one case of invasive infection by ST398 MRSA after ingestion of contaminated food has ever been identified, in a highly immuno-compromised individual in a Dutch hospital (Kluytmans, 2010). Given the large quantities of pork consumed for decades without other cases being identified, the threat to the U.S. population for this pathway appears to be near zero.
  
• 
  Enterotoxicosisby ST398 MRSA. Although staphylococcal enterotoxicosis is a common form of food poisoning caused by Staphylococcus aureus toxins in food, the pig-associated (ST398) strains of MRSA do not produce these toxins, as they lack the enterotoxin genes needed to express them (Argudin et al., 2009). No association between consumption of pig meat and increased risk of such food poisoning has been found (EFSA, 2009). Thus, the threat of food poisoning by ST398 MRSA in pork products also appears to be insignificant or non-existent.
  
• 
  Colonization with ST398 MRSA encountered via the food handling process. Although the probability of this type of colonization appears to be very low for the average consumer, as discussed next, it is somewhat higher for people in meat-handling professions such as butchers and slaughter house workers (Van Cleef et al., 2010; Gilbert et al., 2012). This assessment therefore includes quantification of the risks of colonization arising from food handling.
  

Direct Exposure to Pigs Can Increase Risk of Colonization with MRSA

Several studies have found elevated risk of MRSA colonization in people working closely with pigs and other livestock (Feingold et al., 2012), including farm workers (Smith, TC. et al., 2013), veterinarians (Cuny et al., 2009), and slaughterhouse workers (Van Cleef et al., 2010). However, subsequent infections have not been documented in the U.S. (Limbago, 2010; IARTF, 2011), although there have been some reports of minor soft tissue infections in other countries (Declercq et al., 2008; Denis et al., 2009). Because MRSA colonization has been observed among pig farm workers on MRSA positive farms, even in the U.S.(Smith, TC. et al., 2013), the possibility of subsequent infections cannot be ruled out. These risks, too, are therefore quantified in the following sections.

Hospital Outbreaks of ST398 MRSA are Extremely Rare

The only known death associated with ST398 MRSA occurred in a medically compromised person in a Spanish hospital who sometimes worked on a pig farm (Lozano et al., 2011). There have also been limited reports of person-to-person transmission of ST398 MRSA in Dutch hospitals (van Rijen et al., 2008; Bootsma et al., 2011); however, there has been only one reported instance of an outbreak of ST398 MRSA among patients (Wulf et al., 2007), and the study report did not clearly differentiate between colonization and infection. Because person-to-person transmission has been suggested, the possibility of future outbreaks remains, and is quantified in the following sections.

Community Outbreaks of ST398 Have Not Been Observed

No community outbreaks of ST398 MRSA have been documented, so community outbreaks are probably even less likely than hospital outbreaks. This may be due to the relatively transient nature of most ST398 colonizations. Although some authors have recently argued that living near pigs raises the risks of colonization (Feingold et al., 2012), the statistical methodology behind such ecological associations is suspect, and such reported associations are often meaningless - random spatial trends alone can create statistically significant regression coefficients; and spatial location can be a confounder for spatially distributed exposure and response variables (Cox et al., 2013). Due to its extremely low likelihood, we have not attempted to compute a risk for community outbreaks of ST398 MRSA.

ST398 MRSA is Found in Retail Pork

Turning to food handling hazards, MRSA, including ST398, is occasionally found in retail meats including pork. Recent studies conducted in various parts of the US tested retail pork products for the presence of ST398 MRSA. Table 6.1 summarizes relevant data from these studies.

Table 6.1. ST398 MRSA Prevalence Data from Retail Pork in the U.S.

Study	Samples	ST398 MRSA positive	Rate
(Kelman et al., 2011)	300	1	0.0033
(Waters et al., 2011)	26	1	0.0385
(Davies, 2009)	143	7	0.0490
(Molla et al., 2012)	135	5	0.0370
(O'Brien et al., 2012)	395	26	0.0658
Overall	999	40	0.0400

MRSA Colonization from Food Handling is Possible

A study in the Netherlands of workers who came in frequent contact with raw meat (institutional kitchens and cold meat processing) found no MRSA colonization in 95 participants (de Jonge et al., 2010) although MRSA was present on 2 of 10 (20%) of pork samples from the same locations and 3 of 26 samples of veal and chicken. A recent study of butchers in Hong Kong did find cases of MRSA colonization (Boost et al., 2012). Seventeen of 300 butchers (5.7%) carried MRSA, and ten of these carried strains commonly found in local swine (3.3% of total). A separate study of pig carcasses at markets in Hong Kong found that 16 of 100 carcasses tested positive for swine-associated strains of MRSA (Guardabassi et al., 2009). However, MRSA carriage in the Hong Kong general population is rare (~1.4%) and is largely associated with working in health care (O'Donoghue & Boost, 2004). Therefore, it is plausible that meat handling can be a risk factor for MRSA colonization in Hong Kong, and possibly elsewhere.

In summary, our hazard identification suggests that the major non-negligible risks from pig-associated MRSA arise from colonization, with potential for subsequent infection. The following sections seek to quantify these risks.