Advances in Health Sciences Education

, Volume 14, Supplement 1, pp 51–56

Proposed interventions to decrease the frequency of missed test results

Authors

    • Iowa City Department of Veterans Affairs (VA) Medical Center
    • Division of General Medicine, Department of Internal MedicineUniversity of Iowa Carver College of Medicine
    • Center for Research in the Implementation of Innovative Strategies in Practice (CRIISP) VA HSR&D Center of ExcellenceIowa City Department of Veterans Affairs (VA) Medical Center
  • Peter Cram
    • Division of General Medicine, Department of Internal MedicineUniversity of Iowa Carver College of Medicine
ORIGINAL PAPER

DOI: 10.1007/s10459-009-9180-4

Cite this article as:
Wahls, T.L. & Cram, P. Adv in Health Sci Educ (2009) 14: 51. doi:10.1007/s10459-009-9180-4

Abstract

Numerous studies have identified that delays in diagnosis related to the mishandling of abnormal test results are an import contributor to diagnostic errors. Factors contributing to missed results included organizational factors, provider factors and patient-related factors. At the diagnosis error conference continuing medical education conference in 2008, attendees attended two focus groups dedicated to identification of strategies to lower the frequency of missed results. The recommendations were reviewed and summarized. Improved standardization of the steps involved in the flow of test result information, greater involvement of patients to insure the follow up of test results, and systems re-engineering to improve the management and presentation of data. Focusing the initial interventions on the specific tests which have been identified as high risk for adverse impact on patient outcomes such as tests associated with a possible malignancy or acute coronary syndrome will likely have the most significant impact on clinical outcome and patient satisfaction with care.

Keywords

Diagnosis errorAmbulatory careLaboratory testsRadiology testsComputerized medical recordError

Background

Delays in diagnosis constitute a common medical error and represent a significant threat to patient safety (Kachalia et al. 2006; Holohan et al. 2005; Thomas 2005; Leape et al. 1993, 1991). Available data suggest that problems in test result reporting, often related to the mishandling of abnormal test results, i.e. missed results, are a frequent system contributor to diagnostic delay (Holohan et al. 2005; Singh et al. 2006; Graber et al. 2005). Understanding the scope of the problem is challenging because it is not easy to recognize when an abnormal test result has been missed. Assessment of the frequency of missed test results can be done examining the medical record for evidence of clinician response to an abnormal test. Several authors have examined the incidence of missed results within discrete healthcare systems by focusing on individual tests (e.g., bone density scans, mammograms) and using chart review to examine the record for evidence of clinician awareness (Cram et al. 2005; Edelman 2002; Poon et al. 2004a; Schiff et al. 2000, 2005). Such studies have documented that 2–18% of the clinically significant abnormal clinical pathology results and 23–28% of clinically significant abnormal images lacked evidence of clinician awareness.

Another way to examine the scope of the problem has been to ask clinicians about the frequency in which they encounter patients with missed results. Surveys of primary care clinicians in private practice, academic affiliated private practice, teaching hospitals and Veterans Administration facilities have consistently identified providers who frequently encounter patients with a clinically significant abnormal test result which was lost to follow up (Wahls et al. 2007; Wahls 2007; Poon et al. 2004b).

Because the clinicians in these surveys have reported that nearly a third of the delayed diagnoses were cancers (Wahls 2007; Wahls and Cram 2007), we conducted three separate studies using medical record review to search for evidence of clinician awareness of an abnormal cancer screening test. The setting was a rural Veterans Administration Medical Center with a teaching hospital, resident clinics, staff clinics and community based outpatient clinics, spanning two states. An advanced electronic medical record (EMR) which integrates progress notes, pathology data, imaging results, and appointments was in use during all of the study time periods. In the first study of a cohort of 327 prostate cancer patients in whom a prostate-specific antigen test was obtained prior to cancer diagnosis, we found that 16% had greater than a 6 months lapse prior to evidence of clinician awareness in the medical record (Nepple et al. 2008). In a subsequent study involving 150 colorectal cancer patients, we found 38% had experienced greater than a 6 months delay between abnormal test result and appearance of signs and symptoms suggestive of colorectal cancer (Peleg and Wahls 2008). Notably, more than half of these also had associated patient factors such has homelessness or the presence of major psychiatric diagnoses or medical comorbidities which likely contributed to the delay in diagnosis. In the final study, a cohort of 78 patients in whom an incidental aortic aneurysm greater than 3.5 cm was noted in the radiology report, upon review of the medical record we found that 38% lacked evidence of clinician awareness (Gordon et al. 2009). Again, reasons for delays were often multi-factorial.

In summary, unacceptable delays exist in clinician’s awareness of clinically significant abnormal test results, and these delays are encountered in a variety of clinical settings involving several different types of diagnostic testing. It is also apparent that an advanced electronic medical record, while helpful, is insufficient to prevent a clinically significant abnormal test result from being lost to follow up.

Development of proposed interventions

A small group discussion session was held to consider interventions to decrease the frequency of these delays associated with test result follow up. In our group the focus was improving the management of clinical pathology and other laboratory test results. The process we used to generate the proposed interventions is summarized. The aforementioned research was first presented to the group, followed by the review of a clinical case in which diagnostic delay resulted from delayed appreciation of an abnormal test result. Clinicians were divided into groups of three to four and were asked to develop interventions, which if implemented would decrease the risk of diagnostic delay due to missed results in the future. The small groups had 20 min to develop recommendations which were then presented to the entire group for feedback. Notes from both the individual work groups and the discussion were utilized to generate this article. Approximately 25 clinicians participated in the two focus group sessions. Their recommendations are presented below.

Among all groups there was a consensus that elimination of all missed results would be difficult to achieve. This is in part due to the complex nature of the testing process which involves multiple steps (ordering, completing, reviewing and incorporating findings into the care plan) and multiple individuals across multiple disciplines (clerical, nursing, physician, specialty and primary care providers). Because the scope of the problem is so great and many organizations have resource limitations, focusing quality improvement efforts on specific tests was recommended. Cancer screening studies and tests associated with acute coronary syndrome were highlighted as areas of concern because of the significant morbidity associated with delays in diagnosing these conditions. It was recommended that a local physician survey be conducted to identify which tests should be targeted for performance improvement efforts.

Human error is more likely to occur in the absence of standardized procedures and check lists. Implementation of an electronic medical record (EMR) which integrates clinical notes, appointments, medications, laboratory and imaging results can improve the efficiency and timeliness of the availability of test results. Standardization of human processes associated with test results data management occurs rarely in health care systems with or without electronic medical records. Although electronic medical records decrease the risk of lost paper, without attention to establishing standardization of the electronic and human procedures for handling test results, inconsistent human processes surrounding the handling of test result data lead to predictable data loss. As a result, the group recognized that many systems with an EMR can continue to have missed test results and resultant diagnostic errors.

Therefore increased standardization of the human processes associated with the handling of diagnostic test results was recommended. Achieving such standardization requires buy-in from the physician and nursing leadership and departmental executives within organizations. These process and procedure standardizations should address for both nursing and medical personnel the anticipated hand-offs associated with shift change, scheduled and unscheduled absences, and part-time clinic presence of primary care and specific specialty services. In addition accountability for the completion of the disposition of the actions required on the basis of the test result is needed. In paper-based medical record systems, clinicians typically initial all reviewed test results, but many EMRs lack such documentation of clinician review. Accountability for review of requested tests would be improved by the redesigning the EMR software to implement an electronic signature which documents review of all test results by the ordering clinician.

Interventions focused at the level of the patient were also recommended. Patients can play a key role as a safety net to ensure results are not lost. At the time of testing, patients should be informed of dates and routes (phone call, letter, or face to face) by which notification will occur. Patients should have the option to request direct notification of the test result by the testing center as well as having their test results sent to other individuals or clinicians of their choice. However, the group also believed that such notifications systems should not be automatic. A patient should request the initiation of direct notification, but providers should also have a process to indicate when direct notification is not appropriate. For example patients with declining cognitive capacity or major psychiatric diagnoses such as active psychosis would not be appropriate for an automatic direct notification of test result data.

A number of interventions were proposed to facilitate data management by the providers. Decreasing the volume of test data which the clinician must review after the patient encounter has been completed was considered the most important. Education of providers to decrease the volume of tests which do not provide value, i.e. inappropriate testing, should be ongoing. In addition improved electronic data filters are needed. Only the items which require action should be presented to clinicians. However building such filters will require intensive computer application development and collaboration with clinicians to ensure concordance with clinical practice. Such logic should be able to recognize both when the test is significantly abnormal and when it is, in essence, unchanged from baseline values. Also the EMR should include a notification process which allows clinicians to recognize when specific follow up to abnormal testing has not been completed as requested.

Clinicians need to have test results available while they are with the patient. This minimizes the time and cost of a second interaction for both the patient and the clinicians. This could be accomplished by completing the testing prior to the clinician visit or shortening the time between the request for the test and the availability of the test result. Patients are likely to resist having two appointments, one for testing and one for face time with the clinician however. Increasing the number of tests that are available at point of care testing would therefore be helpful to both the clinician and the patients and should be aggressively implemented.

Many interventions were focused on improving the management of data. Improved system engineering of data management along with clinician involvement for ongoing refinement to the management and presentation of clinical data was universally endorsed. Improvements in data management should also include improvements in systems of accountability. Verification of clinician review of a test results should be documented with an electronic signature (EMR) or a wet signature (paper). A mechanism should also be in place by which departments can monitor who has not completed review of specific test results in a timely fashion to facilitate accountability at the individual and department level.

Addressing a problem with a low incidence rate is generally more effective when tackled from the public health or population perspective. Clinicians are focused on individual patients and do not have the tools or the training to address issues at the level of the population. Few organizations have an individual assigned to focus on population-level concerns. Assigning a specific individual to monitor the follow up to specific high risk tests such as chest x-rays in which a possible mass has been discovered is an example of such a strategy. The most successful organizations will therefore have population focused positions assigned to the follow up of key diagnostic tests and not rely solely on individual clinicians for ensuring proper follow up has occured.

Increased standardization was also identified as being important to reduction of missed results. Establishing standard processes and procedures would decrease variation and the risk of human error. Having a standard process and procedure for hand-off of pending test results between providers is important. This includes the hand-offs between the inpatient and outpatient setting, between specialty and primary care providers and between clinicians within a specialty having periodic shifts in assigned duties typical of teaching hospitals. Standard processes to receive patient requests for their test result information should exist both when the clinic is in session and when it is not. Standard processes to inform patients how they can obtain copies of their medical records including test result information should exist. However if there are no methods to ensure staff and patients are familiar with them, such processes will be inconsistent, not standard.

Conclusion

Abnormal test results lost to follow up occurs across multiple types of tests, with multiple types of clinicians, and in diverse types of clinical settings. An advanced electronic medical record, although helpful in improving timely access to test results, is insufficient to prevent these types of errors. The process by which tests are requested, completed, reviewed and acted upon is complex and numerous opportunities for breakdown in the communication and flow of information can occur. It will be difficult to eliminate these errors because of the complexity of the processes involved.

Recommended strategies include improved standardization of the steps involved in the flow of test result information, greater involvement of patients to insure the follow up of test results, and systems re-engineering to improve the management and presentation of data. Institutions will likely be the most successful with their initial quality improvement efforts if they focus on the specific tests which have been identified as high risk for adverse impact on patient outcomes such as tests associated with a possible malignancy or acute coronary syndrome.

Acknowledgments

This material is the result of work supported with resources and the use of facilities in the Center for Research in the Implementation of Innovative Strategies in Practice (CRIISP) at the Iowa City VA Medical Center.

Copyright information

© Springer Science+Business Media B.V. 2009