Computer Wizardry Not Crucial to Cutting Hospital Errors

KANSAS CITY, MO. – The use of a resident handoff bundle was associated with significant reductions in overall medical errors, regardless of whether a computerized tool linking the electronic medical record was present or not.

Rates of errors, including those that harmed patients, near misses, and minor errors, significantly decreased from 32% to 19% in a unit with the full intervention and from 45% to 25% in a unit with all but the computerized tool. The rate of decline between the two units was not significantly different.

The two general inpatient units at Children’s Hospital Boston had relatively similar patient populations and size, the same resident call system, and the identical preexisting system for verbal and written handoffs, lead author Dr. Amy J. Starmer explained during the plenary session at the Pediatric Hospital Medicine 2011 meeting.

A review of randomly selected sign-out documents also found that several key elements were present significantly more often after implementation of the resident handoff bundle (RHB).

The research intervention featured three components, including communication and handoff skills training for resident physicians, adapted in large part from TeamSTEPPS, a communication and teamwork skills program developed by the Department of Defense and the Agency for Healthcare Research and Quality.

Secondly, the investigators attempted to standardize verbal handoffs by putting a verbal mnemonic in place, restructuring the process so that all residents and interns were brought together as a team and enforcing the process with intermittent evaluations and feedback from the chief resident.

Finally, a computerized handoff tool was introduced in Unit 1 that allowed for pre-text entry of key information fields and automatic importing of information such as demographics, medications, or patient code status, said Dr. Starmer, a pediatrician at Children’s Hospital. Unit 2 used a Microsoft Word document for these written tasks.

Outcomes were measured 3 months pre- and post-intervention using previously standardized error surveillance methods, a review of 107 electronic copies of daily printed sign-outs, and a written miscommunication abstraction tool with 14 key data elements. In all, 529 patient entries were reviewed for Unit 1 and 204 for Unit 2.

Based on the review of Unit 1, certain items were uniformly included preintervention such as name (97%), medical record number (97%), chief complaint (97%), allergies (95%), weight (94%), and patient summary (75%), while others such as medication list (3%), code status (2%), lab results (1%), and vital signs with date (0%) were rarely present, Dr. Starmer said.

All of the aforementioned items, except for chief complaint (99%) and patient summary (95%), were present in 100% of reports postintervention, likely because the data were being automatically imported, she said. Improvements also were observed for more complex items such as to-do lists (29% pre vs. 82% post) and contingency plans (18% vs. 47%), likely because of computer prompting. All comparisons, except chief complaint, were statistically significant (P less than .05).

In Unit 2, significant gains were observed in the inclusion of contingency plans (25% pre vs. 40% post), lab results with date (7% vs. 17%), and vital signs with date (2% vs. 9%) (P less than .05 for all), Dr. Starmer said.

Upward trends were noted for the routinely included items of chief complaint (96% vs. 100%), weight (99% vs. 100%), and patient summary (96% vs. 100%), with slight declines in allergies (94% vs. 91%) and code status (3% vs. 2%). Name and medical record number were included in 100% of records pre- and postintervention.

Limitations of the study include the inability to control for resident learning over time and seasonal variations, its single-site design, and the inability to sustain the use of the verbal mnemonic following the research period, said Dr. Starmer at the meeting, sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association.

She went on to say that these limitations will be addressed in I-PASS (IIPE-PRIS Accelerating Safe Sign-Outs), a multicenter study supported by the Initiative for Innovation in Pediatric Education (IIPE) and the Pediatric Research in Inpatient Settings (PRIS) network.

An attendee asked how the team approach to resident-intern verbal signoffs has been impacted by new resident work hours. Dr. Starmer said there was some concern that the new work rules would affect the amount of time spent with each patient, but that this has not changed.

Dr. Starmer and her coauthors reported no relevant financial relationships.

KANSAS CITY, MO. – The use of a resident handoff bundle was associated with significant reductions in overall medical errors, regardless of whether a computerized tool linking the electronic medical record was present or not.

Rates of errors, including those that harmed patients, near misses, and minor errors, significantly decreased from 32% to 19% in a unit with the full intervention and from 45% to 25% in a unit with all but the computerized tool. The rate of decline between the two units was not significantly different.

The two general inpatient units at Children’s Hospital Boston had relatively similar patient populations and size, the same resident call system, and the identical preexisting system for verbal and written handoffs, lead author Dr. Amy J. Starmer explained during the plenary session at the Pediatric Hospital Medicine 2011 meeting.

A review of randomly selected sign-out documents also found that several key elements were present significantly more often after implementation of the resident handoff bundle (RHB).

The research intervention featured three components, including communication and handoff skills training for resident physicians, adapted in large part from TeamSTEPPS, a communication and teamwork skills program developed by the Department of Defense and the Agency for Healthcare Research and Quality.

Secondly, the investigators attempted to standardize verbal handoffs by putting a verbal mnemonic in place, restructuring the process so that all residents and interns were brought together as a team and enforcing the process with intermittent evaluations and feedback from the chief resident.

Finally, a computerized handoff tool was introduced in Unit 1 that allowed for pre-text entry of key information fields and automatic importing of information such as demographics, medications, or patient code status, said Dr. Starmer, a pediatrician at Children’s Hospital. Unit 2 used a Microsoft Word document for these written tasks.

Outcomes were measured 3 months pre- and post-intervention using previously standardized error surveillance methods, a review of 107 electronic copies of daily printed sign-outs, and a written miscommunication abstraction tool with 14 key data elements. In all, 529 patient entries were reviewed for Unit 1 and 204 for Unit 2.

Based on the review of Unit 1, certain items were uniformly included preintervention such as name (97%), medical record number (97%), chief complaint (97%), allergies (95%), weight (94%), and patient summary (75%), while others such as medication list (3%), code status (2%), lab results (1%), and vital signs with date (0%) were rarely present, Dr. Starmer said.

All of the aforementioned items, except for chief complaint (99%) and patient summary (95%), were present in 100% of reports postintervention, likely because the data were being automatically imported, she said. Improvements also were observed for more complex items such as to-do lists (29% pre vs. 82% post) and contingency plans (18% vs. 47%), likely because of computer prompting. All comparisons, except chief complaint, were statistically significant (P less than .05).

In Unit 2, significant gains were observed in the inclusion of contingency plans (25% pre vs. 40% post), lab results with date (7% vs. 17%), and vital signs with date (2% vs. 9%) (P less than .05 for all), Dr. Starmer said.

Upward trends were noted for the routinely included items of chief complaint (96% vs. 100%), weight (99% vs. 100%), and patient summary (96% vs. 100%), with slight declines in allergies (94% vs. 91%) and code status (3% vs. 2%). Name and medical record number were included in 100% of records pre- and postintervention.

Limitations of the study include the inability to control for resident learning over time and seasonal variations, its single-site design, and the inability to sustain the use of the verbal mnemonic following the research period, said Dr. Starmer at the meeting, sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association.

She went on to say that these limitations will be addressed in I-PASS (IIPE-PRIS Accelerating Safe Sign-Outs), a multicenter study supported by the Initiative for Innovation in Pediatric Education (IIPE) and the Pediatric Research in Inpatient Settings (PRIS) network.

An attendee asked how the team approach to resident-intern verbal signoffs has been impacted by new resident work hours. Dr. Starmer said there was some concern that the new work rules would affect the amount of time spent with each patient, but that this has not changed.

Dr. Starmer and her coauthors reported no relevant financial relationships.

KANSAS CITY, MO. – The use of a resident handoff bundle was associated with significant reductions in overall medical errors, regardless of whether a computerized tool linking the electronic medical record was present or not.

Rates of errors, including those that harmed patients, near misses, and minor errors, significantly decreased from 32% to 19% in a unit with the full intervention and from 45% to 25% in a unit with all but the computerized tool. The rate of decline between the two units was not significantly different.

The two general inpatient units at Children’s Hospital Boston had relatively similar patient populations and size, the same resident call system, and the identical preexisting system for verbal and written handoffs, lead author Dr. Amy J. Starmer explained during the plenary session at the Pediatric Hospital Medicine 2011 meeting.

A review of randomly selected sign-out documents also found that several key elements were present significantly more often after implementation of the resident handoff bundle (RHB).

The research intervention featured three components, including communication and handoff skills training for resident physicians, adapted in large part from TeamSTEPPS, a communication and teamwork skills program developed by the Department of Defense and the Agency for Healthcare Research and Quality.

Secondly, the investigators attempted to standardize verbal handoffs by putting a verbal mnemonic in place, restructuring the process so that all residents and interns were brought together as a team and enforcing the process with intermittent evaluations and feedback from the chief resident.

Finally, a computerized handoff tool was introduced in Unit 1 that allowed for pre-text entry of key information fields and automatic importing of information such as demographics, medications, or patient code status, said Dr. Starmer, a pediatrician at Children’s Hospital. Unit 2 used a Microsoft Word document for these written tasks.

Outcomes were measured 3 months pre- and post-intervention using previously standardized error surveillance methods, a review of 107 electronic copies of daily printed sign-outs, and a written miscommunication abstraction tool with 14 key data elements. In all, 529 patient entries were reviewed for Unit 1 and 204 for Unit 2.

Based on the review of Unit 1, certain items were uniformly included preintervention such as name (97%), medical record number (97%), chief complaint (97%), allergies (95%), weight (94%), and patient summary (75%), while others such as medication list (3%), code status (2%), lab results (1%), and vital signs with date (0%) were rarely present, Dr. Starmer said.

All of the aforementioned items, except for chief complaint (99%) and patient summary (95%), were present in 100% of reports postintervention, likely because the data were being automatically imported, she said. Improvements also were observed for more complex items such as to-do lists (29% pre vs. 82% post) and contingency plans (18% vs. 47%), likely because of computer prompting. All comparisons, except chief complaint, were statistically significant (P less than .05).

In Unit 2, significant gains were observed in the inclusion of contingency plans (25% pre vs. 40% post), lab results with date (7% vs. 17%), and vital signs with date (2% vs. 9%) (P less than .05 for all), Dr. Starmer said.

Upward trends were noted for the routinely included items of chief complaint (96% vs. 100%), weight (99% vs. 100%), and patient summary (96% vs. 100%), with slight declines in allergies (94% vs. 91%) and code status (3% vs. 2%). Name and medical record number were included in 100% of records pre- and postintervention.

Limitations of the study include the inability to control for resident learning over time and seasonal variations, its single-site design, and the inability to sustain the use of the verbal mnemonic following the research period, said Dr. Starmer at the meeting, sponsored by the Society of Hospital Medicine, the American Academy of Pediatrics, and the Academic Pediatric Association.

She went on to say that these limitations will be addressed in I-PASS (IIPE-PRIS Accelerating Safe Sign-Outs), a multicenter study supported by the Initiative for Innovation in Pediatric Education (IIPE) and the Pediatric Research in Inpatient Settings (PRIS) network.

An attendee asked how the team approach to resident-intern verbal signoffs has been impacted by new resident work hours. Dr. Starmer said there was some concern that the new work rules would affect the amount of time spent with each patient, but that this has not changed.

Dr. Starmer and her coauthors reported no relevant financial relationships.