I am nearing retirement and reflecting on my 30-year career in the lab. We are a medium-sized hospital lab. The lab has total laboratory automation, lab information system (LIS) interfaces for most of our instruments, and an electronic medical record (EMR) with computerized physician order entry. Looking back, we don't have fewer patient safety problems, errors, or patient complaints than when I started. In fact, I think it might be worse. What happened? What could I have done differently?
Yours truly,
A Soon-to-be-retired Lab Director

Michael Astion, MD, PhD
Editor-in-Chief Patient Safety Focus

This is what happened. Your frustration level did not go down because even though the lab's error rate is lower, the absolute number of errors is the same or higher as when you started your career. Thirty years ago, the lab was probably performing only two tests, looking at and tasting urine, and it was processing about 150 manual test requisitions per day with 25 employees. Now the lab has 24 employees who perform around 200 different assays, as well as send outs for another 1,000 different tests during the course of a year, and they are processing 1,500 requisitions per day.

In addition, you have been afflicted by harm-causing errors that did not exist when your career started. These errors are due to increased complexity and technology, and include confusing representation of lab results in the EMR and wayward automation that can cause more errors in an hour than lab staff previously made in a month. Moreover, while some of your errors like those caused by mouth pipetting have disappeared, others that futurists predicted would disappear did not because realizing the full potential of robots and information systems has proven to be more difficult that expected.

Reasons Why Patient Safety Issues Have Not Declined

  1. Lower error rates are not associated with a lower number of absolute errors.
  2. New errors have developed in the age of automation and computerization.
  3. Automation and computerization do not eliminate problems with employee performance.
  4. Patient expectations are rising, as is physician disclosure of errors to patients, leading to a higher likelihood of complaints.

Think about it. Despite LIS interfaces, a fair amount of manual data entry still takes place. Even with robotic pipetting and specimen routing, technologists still do a fair amount of manual pipetting, sorting, and transporting. And let's not forget the doctors. They still retrieve lab results from information systems, and they often forget to do so.

Patients also play a role in your frustration. They are enjoying an array of new medications, nutritional supplements, herbs, spices, and teas that could add years to their lives, as well as new interferences to their urine and blood tests. Patients may be feeling better, but you are trying to figure out why their immunoassay results are elevated and never go down, even though no disease is detected.

Patient expectations also are rising as a result of better diagnostic procedures and treatments. In addition, physicians are being urged to disclose errors that previously would have been undisclosed or only partially disclosed to patients. For example, in the past a doctor might have told a patient that a test was going be repeated because the results were inconclusive. Today, the patient is more likely to be told the truth: the specimen was lost in the lab. Disclosing errors is a healthy trend in medicine today, but at the same time, it can lead to increased patient complaints even if the overall number of lab errors is decreasing.

The most important thing that happened during your career was that the nature of human beings did not change. We are likely to make mistakes, especially if placed in an error-prone environment. Automation and disciplined problem solving methods like Lean have helped the lab, but not completely.

10 Ways to Cut Lab Error Rates

  1. Even though stronger interventions such as automation and lab redesign are more difficult and riskier to implement, favor them over weaker ones like telling staff to be more careful.
  2. Mitigate risks associated with new technology by performing more usability testing of all instruments, interfaces, and electronic medical records before implementation. This should include site visits to other labs that have already implemented the technology you are evaluating.
  3. Reduce multi-tasking in the lab, especially by removing phone calls from technologists' responsibilities. Only technologists dedicated to this task should make calls.
  4. Reduce the effect of fatigue on lab errors by avoiding double shifts. Use instead staggered shifts and dedicated night-shift employees.
  5. Deal earlier and more forcefully with human resource issues such as drug- or alcohol-related impairment, incompetence, and disruptiveness.
  6. Assess competency objectively for tasks that require significant skill and are strongly linked to patient safety, such as interpretation of microscope images.
  7. Connect laboratory workers to the patient's experience by arranging for them to have direct contact with patients or time to discuss patients' triumphs and tribulations.
  8. Monitor performance more frequently and fairly because performance improves when employees know their performance is being watched.
  9. Communicate frequently with staff, erring on the side of communicating too much. This is particularly true for major changes that could have tremendous benefit to patient safety if well implemented.
  10. Create work environments that enable the majority of lab staff to work quietly with minimal interruptions.

Taking a Different Approach

Hindsight is 20/20 when it comes to fixing problems. The table (See 10 Ways at left), lists 10 things all lab leaders can do to improve patient safety related to laboratory services. The first one is philosophical, the second involves technology, and the next eight are about people.

If nothing else, remember this: automation is always improving and the use of particular disciplined methods of problem solving wax and wane. Emphasizing the power of people is fundamental to great leadership. A great team can produce great results using a reasonable set of not-so-extraordinary technologies and problem-solving methods, while a mediocre team can do no better than run in place, even with state-of-the-art tools.

I'm sure you've celebrated many successes during your long career. Don't forget them, and enjoy your freedom in retirement.