Читать книгу Social Monitoring for Public Health - Michael J. Paul - Страница 12

CHAPTER 2

Public Health: A Primer

Sandra is a college student. One morning during the fall semester, Sandra wakes up with a fever, cough, and headache. She feels sick enough that she decides to go to her campus’s student health services. At the clinic, a doctor diagnoses Sandra with influenza—the seasonal flu. For a young healthy person with no complications, the treatment is easy enough: drink plenty of fluids, stay in bed, and take ibuprofen or acetaminophen to help with the fever. After a few days, she will hopefully feel better and return to class.

Was it inevitable that Sandra contract the flu or could she have done something to prevent it? For many people, flu is a preventable disease. The seasonal flu shot offers remarkable protection against contracting influenza. The exact rates of immunity vary year to year, but on the whole, it is the single most effective step one can take to prevent a disease that infects tens of millions of Americans, hospitalizes hundreds of thousands, and kills thousands each year.¹ Many universities organize vaccination campaigns at the start of the academic school year, offering free vaccines to students to prevent flu outbreaks. For those who are medically able to receive the vaccination, momentary discomfort and minor side effects are a small price to pay for protection against a potentially deadly virus.

Both of these cases—Sandra receiving treatment for the flu and Sandra receiving a vaccine—are clear examples of healthcare and medicine. However, they are also within the scope of public health.

What exactly is public health and how does it differ from medicine in general? As a rough guide, medicine deals with the interaction of a single doctor and patient: Sandra receiving care from her doctor for influenza. On the other hand, public health focuses on an entire population and asks what can be done to improve the health and prevent disease in mass. How can we inoculate a population to a disease, and track an outbreak as it develops?

How many patients do you need to treat before you are in the domain of public health? After all, developing a new cancer drug can save millions of lives, but we wouldn’t consider it public health. Conversely, an effective public health intervention may only impact a few hundred people. The distinction between the two isn’t a matter of numbers.

Furthermore, the choice of topic alone doesn’t determine if work is considered public health. Some areas of medicine fall almost entirely under the domain of public health, such as controlling environmental pollution to limit negative health outcomes or improving vehicle safety to prevent traffic fatalities. Other areas fall outside of public health research, such as new surgical or radiology techniques. But many domains touch both public health and traditional medicine, including drug overdoses, mental illness, and our example above: infectious disease and vaccinations.

What makes an area a public health topic isn’t the disease or ailment: it’s the types of interventions and goals.

Consider the vaccine example above. If Sandra had visited her doctor earlier in the semester, the physician might have recommended she receive her annual flu shot during her visit. This interaction isn’t in the domain of public health. Furthermore, the research that goes into developing the vaccine isn’t necessarily public health either. Her campus deciding to launch a vaccination campaign with the goal of reaching hundreds or thousands of students, or working with local clinics to ensure they distribute vaccine information—that is public health.

If what defines public health is a set of methods and goals, then what are they? This chapter provides an answer. We will outline the basic goals and principles of public health. Our goal is to provide basic fluency with the field to set the stage for understanding how social media can advance public health understanding.

Before we get into specifics, we should step back to consider the amazing success public health has had so far. Public health efforts have led to safer roads and cars to dramatically reduce traffic deaths, improved workplace safety, reduced pollutants to create safe drinking water, reduced infant mortality, dramatically reduced tobacco use to prevent lung cancer, reduced cavities through water fluoridation, improved our ability to control disease outbreaks, and nearly wiped out several infectious diseases with vaccines.²

Let’s consider the seasonal influenza vaccine as just one example. Many of us are accustomed to our annual flu shot, but it’s worth admiring this marvel of our modern public health system. We often forget just how unique it is to have a treatment that entirely prevents a disease from ever infecting a patient. After all, the first vaccine (for smallpox) was only invented at the turn of the 19th century. In the case of the seasonal flu shot, there are a number of challenging factors.

We’ll focus for a moment on the United States. First, for a variety of reasons, each season’s influenza strain requires its own vaccine. This means that researchers must develop a new vaccine each year. Second, in order to allow enough time to produce millions of doses of vaccines, decisions on what influenza strains should be included in the annual vaccine must be made many months before the start of flu season. Researchers make educated guesses based on currently circulating strains, as well as looking at countries in the southern hemisphere. Since these countries have winter during the U.S. summer, we can gain clues as to what strains may be circulating by looking at their flu season. As an aside, often times the reason that a flu shot is less effective is because unanticipated strains are circulating. Third, rapidly manufacturing safe and effective vaccines for the start of the flu season requires careful coordination between manufacturers and government agencies. Finally, vaccines rolling off the manufacturing line isn’t enough. Health organizations need to decide how many to order, and how to run vaccination campaigns. Should they run lots of advertising early in the season against a possible early seasonal peak, or should they run a longer campaign in anticipation of a late flu season? Will this season be mild, in which case they may not heavily advertise, or will it be a particularly severe season? It is a remarkable feat of the modern public health system that all of this comes together each year and results in tens of millions of Americans receiving a vaccine, preventing numerous infections, hospitalizations and deaths.

With a sense of amazement, let’s proceed to discussing the techniques and goals of public health.

2.1 THE PUBLIC HEALTH CYCLE

Broadly speaking, public health focuses on two distinct goals. First, to monitor and assess the health of a population, including the identification of health problems. Second, to craft health policies to address the identified health problems, including the task of ensuring the population has access to appropriate care. Rather than being distinct goals, they are intertwined. As policies and healthcare practices are revised, public health researchers must reassess the population to understand the effectiveness of the policies and practices and adjust accordingly.

These two goals are reflected in the public health cycle, which consists of ten different components [Harrell and Baker, 1994].³ Figure 2.1 illustrates how these components can be organized around three main activities.

• Assessment: monitoring the health of a population, and identifying and evaluating health issues.

• Policy development: education and development of community partnerships to come up with policies that address the results of assessment.

• Assurance: enforcing policies, providing access to care, and evaluating the results of the policies.

Social monitoring has a role to play in all three activities. It can be used to learn about a population, to help develop partnerships and debate policy, and to provide care. In disease prevention, we can measure infection prevalence in a population (assessment), network to create new partnerships with healthcare organizations to disseminate influenza information (policy development), and provide information on care, for example by sharing links to organizations providing vaccines (assurance). There are numerous examples of social media aiding health communication [Hawn, 2009, Moorhead et al., 2013] and various health interventions [Korda and Itani, 2013]. Many health agencies use social media for broadcasting information [Bartlett and Wurtz, 2015, Harris et al., 2013, Neiger et al., 2013], and doctors use social media to engage patients and the public [Lee et al., 2014b].

Figure 2.1: The three main stages of the public health cycle, along with their respective public health activities, as proposed by the Core Public Health Functions Steering Committee [Harrell and Baker, 1994].

As we said at the end of the previous chapter, this book will focus on assessment, which in the case of social media typically involves the passive monitoring of data to learn about health issues, and their prevalence, in a population. Social media is ideally suited for surveillance: it provides a constant stream of information on a population that can be monitored for topics of relevance to public health. Additionally, monitoring tasks is a more accessible type of research for computer scientists looking to work in public health. The main tools in this domain are data collection and analysis, as opposed to designing interventions or working with patients. That’s not to say that there aren’t many examples of computer scientists working successfully in policy development or assurance. Rather, we find plenty of interesting problems in assessment, and we’re sure you will too!

2.1.1 PUBLIC HEALTH SURVEILLANCE

Public health surveillance concerns the “continuous, systematic collection, analysis and interpretation of health data.”⁴ This includes monitoring for existing identified health concerns as well as discovering new issues. You may also hear the term syndromic surveillance, which is surveillance of a specific syndrome (a set of related symptoms).

Consider infectious disease surveillance, which is one of the largest and most widespread examples of public health surveillance. The United States has a fairly robust national surveillance system for infectious diseases. Perhaps the largest surveillance system is FluView,⁵ the Centers for Disease Control and Prevention’s (CDC) national influenza monitoring system. FluView encompasses several sources of data, including ILINet, a network of thousands of clinics throughout the United States that report weekly statistics on patients presenting with influenza-like illness. These reports, along with virology reports and other sources, make up a weekly CDC report that tracks the rate of influenza infection. A similar process is replicated on the state and local level in many jurisdictions, and many U.S. states produce regular flu reports. Due to its popularity as an application for the use of social media data, we’ll discuss influenza surveillance in detail in Section 5.1.1.

National infectious disease surveillance extends to other notifiable infectious diseases, illnesses in which a physician is required to notify public health authorities of an infection. Examples include measles, ebola, and dengue.⁶ Surveillance also extends to discovering new illnesses. This is how AIDS was originally identified, by the CDC pro-actively investigating unexplained infections [Curran et al., 2011].

Surveillance goes well beyond infectious diseases. Surveillance can identify novel tobacco products [Ayers et al., 2011a, Stanfill et al., 2011] and adverse reactions to medications [Budnitz et al., 2006], both of which have been achieved using social media (see Sections 5.2.2 and 5.4.2, respectively).

2.2 SOURCES OF DATA

Public health depends on data about populations to support its goals. Traditionally, public health draws data from two main sources.

The first is surveys, in particular telephone surveys, which have long been the backbone of public health. There are several, large-scale surveys run on a regular basis (typically annually) that provide a steady supply of public health data. Examples include the Behavioral Risk Factor Surveillance System (BRFSS) and the National Immunization Survey. BRFSS is run annually and collects detailed data from more than 300,000 Americans on a wide variety of public health topics, including access to medical care, mental health, exercise, and tobacco use. Some large-scale surveys rely on in-person interviews, such as the annual U.S. National Survey on Drug Use and Health (NSDUH). Beyond these repeated surveys, many researchers commission onetime telephone or in-person surveys. These can include focus groups, which provide more free-flowing sources of information on public beliefs and attitudes. Online surveys are also growing in popularity due to their low cost, though numerous quality challenges remain [Cook et al., 2000, Dredze et al., 2015, Eysenbach and Wyatt, 2002]. Finally, many private polling companies also conduct health-related phone surveys. For example, Gallup uses phone surveys to measure the well-being of Americans.⁷ A growing thread of work with social media data consider methods for enhancing or replacing traditional survey mechanisms [Benton et al., 2016b].

The second primary source of data come from clinical encounters. The influenza surveillance network described above, ILINet, is the largest such example. Large-scale surveillance networks require significant coordination as they rely on active reporting from clinics. More recently, researchers have turned to automated methods run on electronic medical records that enable scalability and reduce the strain on manual reporters.⁸

While these are the most common data sources for public health, the field has a tradition of seeking new and creative sources of data suited to specific analyses. These include monitoring drug sales and pharmacy records [Heffernan et al., 2004, Magruder et al., 2004] to track gastrointestinal illness [Edge et al., 2004] and use of nicotine replacement therapies [Metzger et al., 2005]. Others have used insurance company billing records to track mammographies [Smith-Bindman et al., 2006] and cardiovascular disease [Lentine et al., 2009]. Some unusual data sources include counting cigarette butt waste in cities [Marah and Novotny, 2011] and estimating community drug abuse by wastewater analysis [Irvine et al., 2011, van Nuijs et al., 2011, Zuccato et al., 2008].

2.2.1 LIMITATIONS OF TRADITIONAL DATA

Monitoring practices that rely on traditional data sources have their advantages and limitations. In general, these methods are well understood and are viewed as reliable, provided they are properly analyzed with biases corrected. Furthermore, many of these data sources go back many years (e.g., annual survey questions), allowing for comparisons over time.

However, we wouldn’t be writing this book if there weren’t disadvantages to traditional methods and thus opportunities for social media data to make improvements. In the case of telephone surveys, they are becoming less accurate over time, as fewer people use landline phones, and the response rate drops [Kempf and Remington, 2007]. This introduces particular bias against low-income and young adults in survey results [Blumberg and Luke, 2007]. Surveys are also expensive to conduct, especially if the survey size is very large or requires in-person interviews [Iannacchione, 2011]. The NSDUH survey mentioned above takes nine months to complete each year.

Clinical records address some of these issues, but are still expensive and complex to set up. Many of the topics covered in surveys do not appear in clinical records, or if they do, they are in unstructured text and thus hard to analyze. Both of these methods can be slow. We cannot measure today’s influenza rate when we do not get clinical records or sentinel site reports more frequently than once a week. Finally, these methods can only cover certain topics, as discussed in Chapter 1. Many areas of public health are understudied because they lack sufficient data to support research.

2.2.2 OPPORTUNITIES FOR SOCIAL MONITORING

These limitations create opportunities for researchers and practitioners to use social media as a data source for learning about health and medicine [Grajales III et al., 2014]. Compared to traditional public health monitoring, social media-based monitoring is fast, cheap, covers a large population, and provides data on topics with little coverage from traditional sources.

One of the most popular social media platforms for health research has been Twitter [Williams et al., 2013], which provides real-time streams of public data, often for free. This type of data creates the potential for real-time health surveillance, which is generally unattainable with traditional methods.

Certainly social media is not a panacea for all problems, and will not replace traditional data sources. We’ll discuss some of these limitations in detail in Chapter 6. However, social media can play a complementary role to traditional monitoring. For example, social media analysis can be used for hypothesis generation [Parker et al., 2015]: rapidly testing out ideas that are not yet worth the time and effort of traditional data collection. The most promising ideas can be forwarded to a more in-depth phase of traditional investigation. Social media can also complement survey data with respect to its demographic coverage. Young adults are overrepresented on Twitter [Duggan et al., 2015] yet underrepresented in telephone surveys, an especially important characteristic for topics like electronic cigarettes and illicit drug use.

A growing chorus of researchers argue that social media will play an important role in public health and epidemiology [Brownstein et al., 2009, Dredze, 2012, Salathé et al., 2012, 2013b]. The U.S. government has taken notice and has started to consider how social media data can aid public health efforts. This has included hosting competitions for building social media-based systems for disease surveillance [Biggerstaff et al., 2016].^{9 10} With social media adoption expanding,¹¹ and with new advancements in technology, social media is likely to have an increasing impact on public health.

¹Flu has such a low mortality rate that many people are surprised to learn that influenza is a leading cause of death in the United States. Since it infects so many people, even a low mortality rate leads to thousands of deaths.

² http://www.who.int/about/role/en/

³ http://www.cdc.gov/nphpsp/essentialservices.html

⁴ http://www.who.int/topics/public_health_surveillance/en/

⁵ http://www.cdc.gov/flu/weekly/

⁶ https://www.cdc.gov/mmwr/mmwr_nd/

⁷ http://www.well-beingindex.com/

⁸ http://www.cdc.gov/ehrmeaningfuluse/syndromic.html

⁹ http://www.nowtrendingchallenge.com/

¹⁰ https://nowtrending.hhs.gov/

¹¹ http://www.pewinternet.org/fact-sheets/social-networking-fact-sheet/