The August 18th issue of the New England Journal of Medicine contained what has to be one of the most robust studies of medical malpractice risk ever conducted. Anupam B. Jena and a bunch of other smarties at Harvard Medical School sifted though over 40,000 records from a major national medical malpractice carrier. The data, collected over a 15-year period, included over 230,000 physician-years of coverage and was large enough to derive specialty-specific risk metrics for 25 specialties. You are encouraged to spend a few minutes reading through their outstanding analysis here. In reading this paper, providers and practice managers can come to a better understanding with respect to several depressing pieces of data, including:
- How likely you are to be sued in a given year
- How likely you are to pay a claim in a given year
- The average size of a claim in your specialty
- Your lifetime risk of facing a claim
- The comparative risk of your specialty relative to other fields of medicine
Given our interest in helping medical groups better manage their medical malpractice risk, we thought it might be useful to slightly extend this work and distill it into a form that would be easily accessible to our uses and prospective clients. Therefore, I reached out to Dr. Jena and requested some of the summary data (which was presented in the paper in graphical form only) and built out a simple model in Excel. The excel model takes Dr. Jena’s data and extrapolates it to calculate the annual risk for a group medical practice. I have placed the data Dr. Jena sent me into the linked Excel workbook so that they can access it if they like (it is in the “Input Parameters” worksheet).
The Math
Though Dr. Jena produced a lot of data, calculating your practices risk is actually fairly simple. To calculate the annual lawsuit risk for your practice, simply calculate the annual risk that an individual physician in your speciality WILL NOT be sued (1-Ri), and raise it to the nth power, where n is the number of physicians in your practice. (1-Ri)^n. This is the annual probability that none of the physicians in your practice will face a lawsuit in a given year. The difference is obviously the risk that you will be sued in a given year. We perform similar calculations to determine the probability that someone in your practice will face a claim. Also, we extrapolate this data out over ten years to show what your overall risk is.
The Model
Below is a brief explanatory video showing how the model works. You can download the Excel model and use it to calculate your practice’s medical malpractice risk. Excel Risk Model.
Over the past several days we have rolled out some new features, most deriving directly from customer requests. These updates include, but are not limited to:
- Provider notification log. When a message is left for the on call provider, we automatically document each attempt at notifying the provider. This audit log is accessible under the actions column for each incoming call.
- Dissertation deterrent. We have added a variable that allows practices to define a maximum message duration. This variable can be used to politely cut off long-winded callers after a reasonable amount of time. The default value, currently applied to all accounts, is 90 seconds. Within a few days you will be able to modify this variable via the On Call Central web interface.
- Updates to call schedule management. We have updated the call schedule management portion of the On Call Central telephony application such that it notifies the user who is currently on call for each call schedule prior to asking the user to put someone new on call. This functionality only applies to accounts with multiple call schedules.
In the past several weeks I’ve been speaking with providers and practice managers frustrated by the quality of after-hours communications. Despite dissatisfaction with their current service, many providers have the impression that a live service is somehow better for patient care. I’ve been skeptical and have been looking for some objective data on this topic. I’ve just found it.
One of the few peer reviewed studies to examine the comparative effectiveness of live –vs- automated communications systems was published about ten years ago by Moss and colleagues in the Department of Medicine at Emory University (where I spent years laboring on my Ph.D.!). The article can be accessed for free here or can be downloaded as a PDF.
The overarching goal of the study was to characterize clinician response times to pages originating from the intensive care unit (ICU) at two university-affiliated hospitals. The study examined a total of 402 pages sent to 166 different physicians when the page was either a) sent directly to the provider or b) through a 3rd party intermediary such as an answering service or a physician’s office. The difference was stark, and most certainly did not support the notion that “live is better.”
Key data points include:
- The median response time for a direct paging system was 2 minutes compared to 9 minutes for indirect systems relying on a 3rd party intermediary.
- Slow responses (≥ 15 min) were observed in 40% of all indirect pages compared with only 8% of direct pages.
- Twenty-five percent of the pages placed through an indirect system (e.g. answering service) were associated with a response time of ≥ 29 min.
Collectively, these data suggest that simple, automated systems are more likely to result in timely responses to patient care issues, particularly in urgent situations.
Today we are pleased to announce the launch of an On Call Central server dedicated to accounts in the State of Alaska. All current On Call Central users in Alaska will now access their accounts at https://your-domain.oncallcentral-ak.com (as opposed to the plain oncallcentral.com domain) and are being issued 907 area code phone numbers.
We Didn’t Do This Alone
We would like to thank several parties for their assistance and technical guidance in navigating the complexities of the Alaska market. Foremost, we would like to thank the people at Tekmate, in particular Alan Zirkle, Buddy Gant, Jared Armstrong, and Shawn Fuller. Not only do these gentlemen know their stuff, but they have been very generous in providing us test machines on which we have been able to optimize our application. We would also like to thank the people at the TelAlaska NOC and the engineers at Alaska Communications Systems for their considerable help. Last, we would like to thank the engineers at Sangoma Technologies, in particular Marc Celsie and Moises Silva, for their assistance in configuring their hardware, which powers core features of this new roll-out.
Why Do This?
The reasons for dedicating resources to Alaska are several, and have become obvious in the past few months. First, Alaska is the only rate center in the United States where major voice over IP providers do not yet issue local phone numbers. We wanted to be able to offer Alaska clients 907 area code numbers, and without establishing a physical infrastructure in Alaska, we would be stuck issuing out of state numbers and 1-800 numbers to clients. Though not a major issue in most cases, we have encountered at least one instance where the lack of a local number made it difficult for clients to dial into their On Call Central account (e.g. when being prompted for a long distance access code at a hospital phone terminal). Second, we were encountering intermittent latency issues that rendered some VoIP calls unusable. This was a fundamental engineering issue resulting from the significant physical distance between Alaska and our Dallas data center. For our Alaska customers, a two party voice call was making a ~10,000 mile round trip between Anchorage and Texas.
What Did We Do?
To resolve these two issues, we embarked on an engineering project that involved, among other things, reconfiguring the telephony portion of the On Call Central code application. Specifically, we moved these customers away from VoIP-based telephony to a system that uses a voice T1 (aka PRI) line. Overall, this entailed what is for us a major investment in time and money, but we expect it to pay off in increased reliability and high quality audio.
