Data Ethics and Healthcare

This past Thursday I had the opportunity to attend a conversation about data ethics hosted by the Kennedy Institute of Ethics, which was very informative. Did you know that the reason women feel cold in rooms is that the room temperature was regulated by the average metabolic rate of MEN?! I didn’t know that either…crazy am I right. I felt that the conversation about data ethics was very much related to healthcare, especially with Apple going into the EMR (Electronic Medical Records) business, so I thought that I would write a bit about data ethics and healthcare.

Before I begin, I’d just like to clarify that EMRs are our medical records but in electronic form. The healthcare industry has mixed opinions about EMRs because they’re a bit annoying. A provider usually has to fill out all of the required information before they can close it, and so it makes their jobs a bit more tedious. Anyways, Apple going into the EMR business is huge because well it’s Apple and it’ll be interesting to see how they integrate their Health app into EMRs.

Okay, now onto the conversation about data ethics.

They had four panelists, Rick Smolan, Cathy O’Neil, Mayra Buvinic, and Chuck Todd. I’ll be writing in the order of the panelists and what their main points were and how it all relates to healthcare.

Rick Smolan

  • His main project concerning data ethics was creating a map that showed where prisoners were going after they were released from incarceration. It turns out these prisoners were going to the same block as before, and so the block was dubbed the million-dollar block because the government is spending millions of dollars moving these people back and from their block.

How Does This Relate to Healthcare?

  • Well if Rick used data to track people’s movements, then data in healthcare can do the same thing, and providers can use that data to see if their patient is getting enough exercise or moving enough in the day. However, the question is whether this is a breach of privacy? Data in healthcare can help providers and insurance companies tremendously, but we must be aware of using the data solely for health reasons and not for profiteering reasons.

Cathy O’Neil

  • Cathy was by far my favorite panelist and had so much to say about data ethics, but the biggest takeaway I took from her was that we are using algorithms to predict success. However, if let these algorithms to dictate our decisions then sometimes they may do more harm than good. Cathy also brought up the point that we might need to an algorithm regulator because we abuse the use of algorithms.

How Does This Relate to Healthcare?

  • Well, say we use algorithms to predict the quality of doctors. If they get a good quality score, then they can keep practicing, but if they get a bad quality score, then they have to go to training or are banned from practicing in that area. If the algorithm is essentially a random process with no mathematical or logical reasoning behind it, then it might tell us to fire the good doctors and keep the bad doctors. Algorithms are used to predict the future because we like to know about the future, but they can be dangerous if there is no basis behind these algorithms and if there’s no regulator behind them.

Mayra Buvinic

  • Mayra didn’t talk much during the conversation, but she had a very important statement about data. Data can lie, and when we believe these lies, then society is in big trouble. People often will manipulate data or not include parts of it so that the data can support their points. This is extremely dangerous because people follow the data, and they will believe it.

How Does This Relate to Healthcare?

  • In healthcare, we often use data to support the growth of ACOs or the expansion of Medicaid/Medicare. And guess who decides whether we expand Medicaid/Medicare? Elected officials. If we manipulate data to convince our elected officials not to expand these programs or decide to block grant them, then millions of people are in danger of losing their healthcare. Data should be used as a supplement, but it should not be abused to hurt others.

Chuck Todd

  • Chuck focused his points on political campaigns, and how they’re primarily run on data. If you’re a reliable Democrat/Republican, then you will most likely have your door knocked on when it campaigns season rolls around. However, if you often switch parties, then your door probably doesn’t get knocked on too much. Chuck believes that campaign practices like these will lead to successful campaigns because they are going after their reliable voters, but it can lead to bad governance because we neglect a huge part of the voters.

How Does This Relate to Healthcare?

  • Say we use data to determine whether we should build another physician facility in the neighborhood. The data they have contains the names of those who regularly go to the doctor. Well, that data might not include those who are uninsured and are afraid to go to the doctor because they don’t want to get turned away. These people might be the one’s who need to see a doctor, but because they don’t go to the doctor, then their names are included in the dataset. What might happen is that we deduce from this data survey that this neighborhood doesn’t need a new physician facility because it seems like everybody has access to doctors. However, we need to be aware of what the data leaves out and how that affects final policy decisions.

Overall, big data is a field that is emerging and can shine some new knowledge about everything. However, we must remember to use data to aid society and not to harm society. We must not take advantage of people and must not use the data to harm others.