Six blind men from Indostan

The American poet, John Godfrey Saxe wrote a poem called "The Blind Men and the Elephant", based upon a much older story about six blind men (from India, at least in one story) who come across an elephant for the very first time and describe the animal based upon their individual perspectives.  The first man touched the elephant's trunk and compared the animal to a snake.  The second man touched the elephant's ear and compared the animal to a large fan.  The third man placed his hand on the elephant's side and described the animal like a stone wall.  I think you get the picture, but the remaining three compared the elephant to a rope (tail), tree (leg), or spear (tusk).  If not, just take a look at the cartoon below (which I believe is by Gary Larson, from The Far Side, but I'm not 100% sure):

As a funny aside, there is also a joke about six blind elephants who come across a man for the very first time.  The first blind elephant accidentally steps on the man and concludes that "Man is flat."  The rest of the elephants all agreed.

The point is that each of us have our own individual perspectives, yet rarely do we get to see everyone's perspective at once.  I am certainly reminded of the High Reliability Organization (HRO) principle of "Reluctance to Simplify" here , as well as the so-called "Rashomon Effect" (and the 2008 movie, "Vantage Point"), which I have discussed before in the past.  As the physicist Werner Heisenberg (most famous for the "Heisenberg uncertainty principle") eloquently stated, "We have to remember that what we observe is not nature in itself, but nature exposed to our method of questioning."

Health care organizations that aspire to be High Reliability Organizations (HROs) analyze adverse events and so-called near misses in order to maximize the information that can be learned in order to prevent these events from happening again in the future (see "Preoccupation with Failure").  Many, if not most, hospitals have invested in web-based incident reporting systems that allow staff members to report their concerns about patient and staff safety.  For example, providers will report if a patient receives the wrong dose of medication or even the wrong medication.  These events will be reviewed by both local and hospital safety leaders to determine what factors were responsible for the error.  Safety leaders can also monitor the incident reporting system for certain trends or patterns, which then often prompt more formal, systemic corrective measures.

With this in mind, I was interested to read a study published in 2010 by a team at Brigham and Women's Hospital in Boston ("Integrating incident data from five reporting systems to assess patient safety: Making sense of the elephant") and the accompanying editorial, "The elephant of patient safety: what you see depends on how you look"in the same issue of the Joint Commission Journal on Quality and Patient Safety.  These investigators compared the nature of safety reports obtained through five different commonly utilized systems - voluntary incident reporting systems (as described in the paragraph above), patient and family complaints, reports to risk management, medical malpractice claims, and executive walk rounds.  They collected data for a 22 month period (2004-2006), though they included 10 years worth of malpractice claims data.

Here is the breakdown of events that they reviewed and classified (using a system that they developed and validated):

1. Incident Reporting - 8,616 reports

2. Reports to Risk Management - 1,003 reports

3. Patient/Family Complaints - 4,722 reports

4. Executive Walk Rounds - 61 walk rounds

5. Malpractice Claims - 322 claims (over 10 years, rather than 22 months)

Importantly, the individuals who completed each report was very different.  By far, bedside nurses completed most of the reports submitted to the incident reporting system (only 2.5% of these reports were submitted by physicians - this is very similar to my anecdotal experience), while over 50% of the risk management reports were reported by physicians.  As a result, the categories of incidents reported were very different between the five systems.  

1. Incident Reporting: Patient identification issues (e.g. wrong patient or wrong medication), Falls, and Medication errors

2. Reports to Risk Management: Issues around technical skills, patient and family behavior, administrative issues, or concerns about clinical judgement/decisionmaking

3. Patient/Family Complaints: Communication, Ancillary services (patient transport, kitchen, environmental serivces), administrative issues (e.g. scheduling or care coordination issues)

4. Executive Walk Rounds: Equipment problems, Electronic Medical Record, Infrastructure or resources (staffing, available supplies)

5. Malpractice Claims: Clinical judgement and/or decisionmaking, Communication, Technical skills, problems with medical records

In other words, any hospital that relies upon only one or even a couple of these incident reporting methods will, just like the six men of Indostan, not have the complete picture!  Each of the five systems capture different issues, and in many cases, these issues would not be identified in any of the other systems.  Similarly, any study reporting patient safety incidents using only one of these different systems should be viewed with some level of skepticism.