Improving patient safety and its reporting system
By Christopher Guadagnino, Ph.D.
Michael Doering is the executive director of the Pennsylvania Patient Safety Authority, which recently released its latest Annual Report.
PND: What progress has been made in improving patient safety in Pennsylvania?
MD: There has been significant progress among different entities in Pennsylvania and by facilities and providers themselves. In June 2004 the Patient Safety Authority began implementation of the Pennsylvania Patient Safety Reporting System (PA-PSRS), and we receive about 210,000 reports each year of incidents and serious events. A serious event is when there’s harm to a patient, and an incident is when there is no harm to a patient. About 96 percent of what we receive are incidents, where there is no harm. Our analysts take that information, review it and put out the Patient Safety Advisory, which is a quarterly publication. We may have direct contact with facilities. We may conduct special studies based on information that we see. We put together patient safety toolkits for folks to use. That’s primarily what we had done in the first three years.
Each issue of the Patient Safety Advisory has about eight to ten articles that have to do with different patient safety events that we see from the reports that are submitted to us. We provide a narrative of what some of these reports are, summarize what the literature says about the types of events, and provide some guidance to the facilities. We do an annual survey of patient survey officers, and 68 percent of those participating in the 2007 survey reported making or planning to make changes based on an Advisory article, and that’s up from 63 percent the previous year. Many of those facilities are implementing numerous pieces of guidance that come from the advisories. Of the 103 hospitals that responded, they said they made 364 changes. So, the information we’re putting out is being used. In 2006 we were awarded the John Eisenberg award by the Joint Commission and the National Quality Forum. That’s something that, among peers in patient safety, is a fairly prestigious national award. We aren’t just sitting back on the laurels of the award, though.
PND: Do you have any data as to whether the advisories drive more reporting of certain events or incidents?
MD: I think it does drive more reporting for a couple of reasons. One, I think facilities have been very responsive to the information that we send out through the Patient Safety Advisory and, in terms of the quality and content of information that we put out, we get graded very highly by the patient safety officers. I think that has led to more reporting. But also, whenever you focus a spotlight on a particular event, sometimes we actually see more reporting after that because facilities begin paying more attention to that type of event. They put processes in place, and they can identify more of the type of activity that can then be reported to us.
PND: Has the volume of events and incidents changed over the past few years?
MD: We’ve seen a significant increase in the volume of reports that we’ve received. That increase has primarily been with incidents, where there is no harm to a patient, and I attribute this to an increased understanding and awareness of reporting, the culture of reporting within facilities. For serious events, where there is harm to a patient, we have not seen an increase in the number or percentage of reports that we receive. What we would like to see, in the end, is the reduction of serious events where someone is harmed.
PND: Is there evidence to suggest that patient safety is actually improving in Pennsylvania.?
MD: It’s difficult to be able to put a number on patient safety because no-one has a baseline that says this is what patient safety is in Pennsylvania. There are no real statistics based on the types of events that I would consider to be patient safety events. So we have to, in some cases, rely on narrative. We have to rely on knowing that there are process improvements being made by facilities. Unfortunately, the data that we have are reports of events happening, and we don’t necessarily have a firm denominator for those events, nor can we be sure that we have all of the events that are actually taking place. We know that there are interventions that are in place. If the number of incidents – where there is no harm to a patient – goes up, we cannot be certain that that’s because the facility is less safe, or a group of facilities is less safe. It just may mean that more incidents are being reported. We’re not going to kid ourselves and think that all incidents are being reported in Pennsylvania facilities.
We just had a retreat with our board to take a look at how to measure patient safety in Pennsylvania. Right now we don’t feel comfortable, with the information that we have, to be able to say, “Here’s what patient safety was at one point, in terms of a statistic, and here’s where it is now.” I don’t think that there are any practitioners in this field who would say that the data out there are perfect to be able to do that. So, we’re going to attempt to put together some sort of an algorithm that we can use to represent patient safety in Pennsylvania.
PND: Are there competing frameworks of definitions for patient safety, or is it a matter of the number of events and incidents going down from some baseline?
MD: It’s not just the number of what’s reported to us going down because, again, I don’t think that we receive everything that is out there. So, the fact that a number would go up doesn’t necessarily mean we’re less safe, and a number going down doesn’t necessarily mean we are safer, in terms of what’s being reported to us. We do have to try to look at a variety of things, including the culture of safety within facilities. That’s something that various organizations are attempting to measure. In some cases we can look at outcome information. In some cases we can look at administrative information. We can also look at things like wrong site surgery – that’s something we believe has a high degree of reporting; we don’t think a lot of those are missed. Where there’s an event that we think is being highly reported and we can have some sort of program to try to reduce it, that’s something that we believe can become part of the algorithm. We would like in our next annual report to be able to define how we’re going to do it and attempt to form a baseline.
PND: What have been the limitations and obstacles to knowing, with confidence, that the number of reported events and incidents correspond to the entire universe of reportable events?
MD: As we pointed out in our 2007 Annual Report, there is a discrepancy between the rates of reporting by facilities – some facilities are reporting up to 50 incidents and serious events per 1,000 patient days and others are reporting much less. There is a variety of possible reasons for that, including the culture of reporting within a facility and how many systems are in place to be able to identify potential errors. But I think a lot of it has to go back to Act 13: the definition of what is reportable as a serious event. Frankly, there are some ambiguous terms, including the word “unanticipated.” Whether or not an event is anticipated or unanticipated largely defines whether it is reportable or not. There’s also the language for a serious event that it has to be something where the patient is injured and additional health care services are provided to the patient. Folks read “additional health care services” in different ways. So, the facilities themselves have interpreted these definitions in different ways, and I think that accounts for a significant portion of the differences in reporting between facilities.
PND: How much variation is there in volume of reported events and incidents?
MD: There’s a significant amount of difference. There are outliers, of course, in all of this, but I would say that the top quartile of hospitals reports between 36 and 300 incidents and serious events per thousand days and the median is around 20. If we go down to the lowest quartile, there are a few hospitals that are zero, and others well into the single digits. A large urban teaching hospital may report five serious events per thousand patient days while another similar hospital might report less than 0.1. So, the difference is not because of the type of hospital.
PND: Your organization has been providing educational guidance to attempt to provide clarity. Why have those efforts failed to substantially decrease reporting variability?
MD: I don’t know that it has failed to decrease variability from where it was before because we never measured it before, but it obviously hasn’t taken it to where it should be. One of the reasons is that the Patient Safety Authority is an educational and training organization. That is our goal. We don’t have any regulatory authority. The Department of Health is the regulator for Act 13. We did put out a memo that said just because something is on the patient consent form doesn’t mean that it is an anticipated event. However, we believe there are facilities that still say, “Well if it’s on the patient consent form, then it was obviously anticipated.” I have been in contact with the Department of Health and we’re about to begin an effort with them to try to define some of these terms better and provide more concrete guidance to the facilities. And then we need to make sure that the surveyors from the Department of Health are trained and are all applying this guidance in the same manner when they do the licensure reviews of the facilities.
We want to be able to understand the characteristics of facilities that are reporting a lot – and again, this doesn’t mean necessarily that they are less safe facilities, because most of what they’re reporting are incidents where there is no harm – because those facilities appear to have a comfort about reporting that information. And they’re not just reporting it to us, they’re reporting it to themselves, because PA-PSRS allows them to review their data and run a variety of reports that most of them will send to their patient safety committees and to their boards of trustees. We’re going to do a survey of facilities that are high reporters and facilities that are low reporters and try to figure out if there are any common characteristics among each group. We’re going to do this anonymously; our goal is not to identify who the high or the low reporters are, but just to be able to understand things like whether the patient safety officer or organization has access to the executive management within the facility, how involved are the chiefs of the various specialties, how involved are the physicians, what is the culture of disclosure at these types of facilities. We think we may be able to gain some valuable insight we can pass along to other facilities.
PND: What other new initiatives do you have planned?
MD: Our board in 2007 said we want a strategic plan going forward. We looked at the work we’d done the last couple of years and it has been focused on getting the reporting system in place, being able to do some analysis, conducting some training and providing guidance through the advisories. We wanted to expand beyond that into more collaboration and into more education and training. We came up with 11 initiatives. One is training and education for hospital boards of trustees, which we are piloting this fall together with the Hospital & Healthsystem Association of Pennsylvania. The Patient Safety Authority is also going to hire a director of educational programs and hire up to six patient safety liaisons, the job of whom is to get out into the facilities to help them to understand what types of education or training we can help them with; as well as to be able to share process improvements, procedures and policy changes between facilities; and to be a resource for those facilities in various regions.
Another initiative is the Patient Safety Knowledge Exchange (PasSKEy). There are so many people undertaking quality patient safety efforts in hospitals and ambulatory surgical facilities within Pennsylvania and you don’t want to have to re-invent the wheel at all of these different facilities. The goal of PasSKEy is to provide an online community for the patient safety officers to be able to discuss different patient safety events and to be able to post and share different policies and procedures that folks have implemented in their particular facility that may be of value and help to others. If something works in a particular area in one facility, it would be great to be able to share that information across many different facilities.
Another new initiative is nursing home reporting, which grew from Act 52’s mandate to collect information on hospital-acquired infections (HAIs). That’s a pretty big job because we have 500 facilities in PA-PSRS right now, and there are an additional 700 or 750 nursing homes, so it’s a significant amount of new facilities that would be reporting to us, and also to the Department of Health. Along with the Department of Health, we have defined a draft of what is reportable and we’re in the comment period right now for that. We were charged with setting up an HAI Advisory Panel in the Commonwealth and we’ve done that: we have a great group of 15 clinicians from around the state who are experts in this field. Five of them were appointed to a long term care committee on reporting, and they helped establish what would be reportable. We’re also working with the Advisory Panel in terms of what type of training we should have for people in facilities. We don’t believe you should just tell them to report and then wait and see what comes in. We want to make it as understandable and easy as possible for them, and we will hold training sessions around the state to help these folks do reporting.