qu- ct 9

452.full.pdf

Home >English homework help >Article writing homework help >qu- ct 9

ORIGINAL RESEARCH

Reporting and Using Near-miss Events to Improve Patient Safety in Diverse Primary Care Practices: A Collaborative Approach to Learning from Our Mistakes Steven Crane, MD, Philip D. Sloane, MD, Nancy Elder, MD, Lauren Cohen, MA, Natascha Laughtenschlaeger, MD, Kathleen Walsh, BA, and Sheryl Zimmerman, PhD

Purpose: Near-miss events represent an opportunity to identify and correct errors that jeopardize pa- tient safety. This study was undertaken to assess the feasibility of a near-miss reporting system in pri- mary care practices and to describe initial reports and practice responses to them.

Methods: We implemented a web-based, anonymous near-miss reporting system into 7 diverse practices, collecting and categorizing all reports. At the end of the study period, we interviewed practice leaders to de- termine how the near-miss reports were used for quality improvement (QI) in each practice.

Results: All 7 practices successfully implemented the system, reporting 632 near-miss events in 9 months and initiating 32 QI projects based on the reports. The most frequent events reported were breakdowns in office processes (47.3%); of these, filing errors were most common, with 38% of these errors judged by external coders to be high risk for an adverse event. Electronic medical records were the primary or secondary cause of the error in 7.8% and 14.4% of reported cases, respectively. The pat- tern of near-miss events across these diverse practices was similar.

Conclusions: Anonymous near-miss reporting can be successfully implemented in primary care prac- tices. Near-miss events occur frequently in office practice, primarily involve administrative and commu- nication problems, and can pose a serious threat to patient safety; they can, however, be used by prac- tice leaders to implement QI changes. ( J Am Board Fam Med 2015;28:452– 460.)

Keywords: Medical Errors, Physician’s Practice Patterns, Practice Management, Quality of Health Care

Near-miss events, or errors that are corrected before a patient is harmed, represent an opportunity to iden- tify and correct flaws that jeopardize patient safety. Because more than half of all medical ambulatory visits occur in primary care, improved attention to near-miss events could markedly improve overall pa- tient safety.1 Others have demonstrated that error- and event-reporting systems can be implemented in primary care; however, these rarely focus on near

misses or the coordination of near-miss reports with quality improvement (QI).2–9

Barriers to reporting events include the addi- tional workload burden, concern over punitive action, lack of confidence that positive change will result, and psychological barriers to admit- ting an error.10 –14 Anonymous reporting systems may increase the number of error reports and reduce concerns about punitive actions but might

This article was externally peer reviewed. Submitted 2 February 2014; revised 30 March 2015; ac-

cepted 14 April 2015. From the Mountain Area Health Education Center, Ashe-

ville, NC (SC, NL, KW); the Cecil G. Sheps Center for Health Services Research (PDS, LC, SZ), and Department of Family Medicine and School of Medicine (PS), and School of Social Work (SZ), University of North Carolina—Chapel Hill, Cha- pel Hill and the University of Cincinnati, Cincinnati, OH (NE).

Funding: This study was funded by grant no. PS/R21 HS19558-01 from the US Agency for HealthCare Research and Quality (“Ambulatory Near Miss Reporting and Track- ing to Improve Patient Safety,” Steven Crane, MD, Princi- pal Investigator, September 2010 –2011).

Conflict of interest: none declared. Corresponding author: Steven D. Crane, Mountain Area

Health Education Center, University of North Carolina— Chapel Hill, 121 Henderson Rd, Asheville, NC 28803 (E-mail: steven.crane@msj.org).

452 JABFM July–August 2015 Vol. 28 No. 4 http://www.jabfm.org

mailto:steven.crane@msj.org

reduce the detail of the events.15,16 There is value in including all office staff in a reporting system, but this strategy may require frequent reminders to keep reporting volumes from dwindling.17

While errors occur frequently in primary care, few seem to result in significant harm to patients, consistent with the “near-miss” nature of many of these errors.18 –22 Nevertheless, given the volume of ambulatory visits, even these relatively infre- quent adverse events may be associated with a sub- stantial portion of inpatient admissions and other patient harm.23 A systematic approach to identify and correct near-miss events in primary care could be an important strategy to improve patient safety.

To demonstrate that such a system can be suc- cessfully adopted by a broad range of primary care practices, we designed and implemented an anon- ymous, practice-wide near-miss reporting and im- provement tracking system in 7 diverse primary care medical practices. Our goals were to assess the feasibility of regular reporting, better understand the types of near-miss events that occur in ambu- latory practices, and observe how medical practices use near-miss reports to initiate QI changes.

Methods Participants We recruited 7 diverse practices in western North Carolina to participate in this 1-year study. Practices included 2 family medicine residency practices, a fed- erally qualified health center, a county-owned health department, and 3 private practices (2 family medi- cine, 1 pediatrics). Together the study practices em- ployed more than 70 medical providers and 200 clin- ical support staff, provided �2000 office visits per month, and represented the full scope of primary care services (pediatric, geriatric, adult, and obstetric care) in both rural and urban settings. All but 1 used elec- tronic medical records. Table 1 summarizes descrip- tive data on these practices.

Near-miss Reporting System Our operational definition of a near-miss event was “an event/situation in which a negative outcome could have occurred but did not, either by chance or because the problem was identified and cor- rected before a negative outcome occurred.”24 All staff members were invited to anonymously report near-miss events using an online form that had been adapted from previous studies and field tested,

Table 1. Description of the Study Practices

Type of practice n (%)

Private 1 (14.3)

Part of a hospital system or other health system

3 (42.9)

Community health center 1 (14.3)

County health department 1 (14.3)

Residency program 1 (14.3)

Primary medical specialty represented

n (%)

Family medicine 6 (85.7)

Pediatrics 1 (14.3)

Number of Providers (full-time equivalents)

Mean (SD); Range

Physicians 3.9 (2.8); 1–8.75

Nurse practitioner or physician assistants

2.8 (2.3); 0.5–5.6

Physicians-in-training (residents)

5.2 (10.2); 0–27

Services provided n (%)

Pediatric care 7 (100)

Obstetric care 3 (42.9)

Geriatric care 5 (71.4)

Number of Patient Encounters Per Year

Mean (SD); Range

Outpatient medical visits 22,589 (11,258.6); 10,000–36,000

Inpatient visits 3,433 (2,528.6); 0–5,688

Obstetric deliveries 230 (122.4); 0–315

Behavioral health visits 2,584 (1,835.0); 0–4,637

Nursing home visits 1,077 (748.5); 0–2,000

Home visits 41.6 (74.6); 0–185

Predominant medical record system for office visits

n (%)

Paper 1 (14.3)

Electronic 6 (85.7)

Percentage of Annual Patient Visits, by Age Category

Mean (SD); Range

� 18 years 52 (44.0); 10–100

18–64 years 47 (31.2); 0–75

� 65 years 16 (17.8); 0–50

Percentage of Annual Patient Visits, by Payer Status

Mean (SD); Range

Private insurance 27 (23.4); 0–70

HMO 3 (8.0); 0–22

Medicare 26 (23.2); 0–60

Medicaid 22 (14.5); 0–45

Self-pay 11 (15.1); 0–45

Charity 11 (23.1); 0–63

Percentage of Annual Visits by Patient Race/Ethnicity

Mean (SD); Range

White 88 (9.1); 75–95

African American 9 (6.2); 4–20

American Indian 0 (0.8); 0–2

Asian 1 (1.8); 0–5

Other 2 (3.3); 0–9

Hispanic or Latino Ethnicity 16 (18.8); 1–40

SD, standard deviation.

doi: 10.3122/jabfm.2015.04.140050 Using Near-Miss Events to Improve Patient Safety 453

with an average completion time of 2 minutes per report (See Appendix).25 The online form did not include any patient identifiers, was available elec- tronically from any Internet-enabled computer, and stored reports on a central computer in an encrypted format. Staff attended a standardized, 1-hour orientation and during the study period received an automated E-mail message every 2 weeks inviting them to report any near-miss event they could recall from the previous 2 weeks. Project participation was phased in over 2 months from September 15 to November 30, 2010, and data collection was terminated at the end of June 2011; thus, the project period last 7 to 9 months, depend- ing on practice site.

Near-miss Event Reports Before being forwarded to the project’s central computer, each near-miss report was reviewed by a designated individual in the practice (usually the medical director), who (1) excluded from the study any events that were adverse events causing patient harm; (2) ensured the absence of patient-identify- ing data in the responses forwarded for analysis; and (3) reviewed the incident for possible initiation of QI efforts in the practice. There was no attempt to standardize across practices which near-miss re- ports would be assigned for QI; during the struc- tured interview process after the study, medical directors and practice administrators reported that they concentrated on events that seemed to be likely to recur, would have potential serious conse- quences if harm reached the patient, and seemed to be in their control to change.

QI from Event Reports Initiation of QI around near misses was encouraged as part of the project. At the time of enrollment in the study, the 7 practices had significant differences in how they approached performance improve- ment. Several had robust QI teams in place, whereas others reported no formal performance improvement processes; none had incorporated near-miss reporting into the QI process. As part of study orientation, practice leaders each received a short orientation including a brief overview of how to initiate a Plan Do Study Act (PDSA) cycle and how to use the PDSA tracking software that was included in the near-miss system. After 3 months of successful reporting, each practice was expected to initiate at least 1 improvement process based on the

near-miss reports from the practice. All near-miss reports within each practice were reviewed every 2 months during a QI committee meeting. At the end of the project period, leaders from each practice participated in a structured group interview to gather additional information about how they ac- tually responded to the information contained in the near-miss reports.

Each practice was reimbursed $5,000 for iden- tifying a core implementation team, participating in planning meetings and the all-staff orientation, and completing the baseline survey information. An additional $1,500 per month was given to each practice when they reported at least 10 near-miss events and identified at least 1 near-miss event to remediate and track. Staff themselves did not re- ceive any direct monetary inducement to submit reports, but several practices introduced small team-based rewards if the practice overall met the monthly reporting target. During the structured group interviews after the reporting period, there were no reports that staff felt pressured to report.

Data Analysis After standardized training, the narrative portions of the near-miss error reports were coded by a team of 6 physician coders using a published taxonomy of am- bulatory care errors.26 For each report, the primary error was defined as “the breakdown in process, or knowledge/skill deficit that led to the reported prob- lem.” In addition, up to 4 associated or “cascade” errors and up to 4 contributing factors and possible preventive measures also were coded using the same taxonomy. The coders also provided their own subjective ratings, on a scale of 0 to 100, of the potential seriousness of the near-miss event, where 0 indicated “not very serious” and 100 indicated “extremely serious.” They also rated the likelihood of harm and potential cost to the patient had the error actually occurred, as well as the estimated cost to the practice to remedy the system problem iden- tified in the near-miss report, all on a 3-point scale, where 0 � “none/minimal,” 1 � “some,” and 2 � “a lot.” Before coding, study leadership and coders met to achieve a common understanding about what would classify as “very serious” or “a lot of harm or cost.” To ensure reliability of the coding and rating, 10% of the reports were coded inde- pendently by a second coder without knowledge of the first coder’s results. Coder agreement was 70%

454 JABFM July–August 2015 Vol. 28 No. 4 http://www.jabfm.org

at the finest level of detail (3 levels in the 5-level taxonomy) and 87% at 2 levels of detail.

Quantitative data were analyzed using SAS 9.1 software (SAS, Inc., Cary, NC). Continuous data are reported using means and standard deviations (SDs), whereas categorical data are reported as fre- quencies and percentages. The study protocol was reviewed by and received institutional review board approval from Margaret R. Pardee Hospital.

Results A total of 632 near misses were reported by the 7 practices. The most common categories of reported near-miss events, overall and by practice, are summa- rized in Table 2. The most common types of errors were breakdowns in office processes (47.3%), such as filing (25.3%), chart data entry errors (15.0%), prob- lems with patient flow (2.2%), and problems with appointments and referrals (4.8%). The second most common category of errors was in ordering (6.2%), implementing (7.1%), or reporting the results of (12.2%) investigations, representing 25.5% of all near-miss reports. The pattern of near-miss events was similar across practices. Errors involving clinical knowledge or performance represented a very small percentage of errors (1.9%).

Table 3 reports coder ratings of near-miss severity, likelihood of an adverse event (AE) if the near miss had not been identified, the potential financial costs if the near-miss event had resulted in an AE, and the estimated cost to the practice to remedy the problem. Filing errors, the most common single near miss reported, had a mean severity rating of 51.8 (SD, 30.7), with 23.8% (n � 38) of these errors judged to be at high risk for leading to an AE had the error not been identified. Among all error types, those related to reporting investigations were rated as potentially most serious, with a mean severity score of 72.0 (SD, 28.3). Errors involving ordering medication and treatments (ordering, dispensing, or implementing) represented only 14% of near-miss reports but were rated as the second most severe (mean range, 59.1– 63.0; SD range, 29.3–31.0).

Practices reported that 14.4% of the errors were secondarily attributable to the electronic medical re- cord (EMR), including 21.9% of the filing errors, an example of which was an EMR interface that did not deliver the results of an important test to the ordering provider and resulted in a delay in addressing the test results. The EMR also was implicated in 40% of

ordering medication or treatment errors and 4.3% of communication with other health care providers shar- ing patient care errors. These computer-related er- rors had the third highest mean severity rating (mean, 59.2; SD, 25.2). The EMR was the perceived primary cause of 49 of these errors (7.8% of the total sample).

By the end of the study period, each of the prac- tices had initiated at least 1 practice improvement process directly tied to the near-miss reports. Table 4 summarizes these practice improvement efforts.

Discussion This study reports the results of the successful introduction of near-miss reporting in 7 primary care practices, each of which generated a substan- tial number and broad array of events and initiated performance improvement activities as a result.

The most frequent near-miss events recorded involved relatively mundane office processes such as charting data, filing, and computer operation, which is consistent with previous reports.26 Some- what surprisingly, however, our data showed that administrative errors were frequently considered to carry with them the potential to lead to significant patient AEs, which supports our approach of en- couraging all office staff to be involved in near-miss reporting. The events judged to be associated with the highest potential cost were those involving dis- pensing medication or implementing treatment (30% judged to involve “a lot” of potential cost) and handling test results (22% judged as “a lot”).

EMRs were directly linked to 14% of near-miss events, including 40% of the errors related to pre- scribing. Among the filing, data retrieval, and pre- scribing errors, 21.9%, 28.4%, and 40% of near- miss events, respectively, were attributed to EMR use. This finding reflects what others have found: While EMRs can reduce errors, they can also cause errors.27–29 Additional study of this important find- ing is needed to redesign EMRs to reduce error rates.

Participating practices seemed to have used the data generated from these near-miss reports to im- plement meaningful practice changes and improve- ments. Each initiated at least 1 Continuous Quailty Improvement (CQI) project as a result of the study, and each identified at least 1 important safety im- provement they made as a result of a near-miss report. Our interview data suggest that practice leaders found that immediate action or rapid PDSA

doi: 10.3122/jabfm.2015.04.140050 Using Near-Miss Events to Improve Patient Safety 455

Ta bl

e 2.

Fr eq

ue nc

y of

N ea

r M

is s

Ev en

ts O

ve ra

ll an

d by

Pa rt

ic ip

at in

g Pr

im ar

y Ca

re Pr

ac ti

N ea

r M

is s

E ve

n t

N u

m b

er o

f N

ea r

M is

s E

ve n

ts (%

o f

R ep

o rt

s, b

y S

it e)

A ll

P ra

ct ic

es (n

� 63

2) P

ra ct

ic e

A (n

� 43

) P

ra ct

ic e

B (n

� 17

7) P

ra ct

ic e

C (n

� 80

) P

ra ct

ic e

D (n

� 13

9) P

ra ct

ic e

E (n

� 69

) P

ra ct

ic e

F (n

� 43

) P

ra ct

ic e

G (n

� 81

)

O ffi

ce P

ro ce

ss P

ro bl

em F

ili ng

a 16

0 (2

5. 3)

7 (1

6. 3)

72 (4

0. 7)

25 (3

1. 3)

18 (1

3. 0)

7 (1

0. 1)

15 (3

4. 9)

16 (1

9. 8)

C ha

rt da

ta a

95 (1

5. 0)

6 (1

4. 0)

20 (1

1. 3)

12 (1

5. 0)

14 (1

0. 1)

10 (1

4. 5)

4 (9

.3 )

29 (3

5. 8)

P at

ie nt

fl ow

14 (2

.2 )

1 (2

.3 )

2 (1

.1 )

3 (3

.8 )

5 (3

.6 )

1 (1

.5 )

1 (2

.3 )

1 (1

.2 )

A pp

oi nt

m en

t or

re fe

rr al

30 (4

.8 )

1 (2

.3 )

4 (2

.3 )

3 (3

.8 )

14 (1

0. 1)

7 (1

0. 1)

1 (2

.3 )

— E

qu ip

m en

t or

B ui

ld in

g P

ro bl

em E

qu ip

m en

t an

d ph

ys ic

al bu

ild in

g/ su

rr ou

nd in

gs /p

ra ct

ic e

si te

8 (1

.3 )

— 2

(1 .1

) 1

(1 .3

) 3

(2 .2

) 2

(2 .9

) —

—

O th

er sp

ec ifi

c pr

ob le

m s

w it

h co

m pu

te r

8 (1

.3 )

— 1

(0 .6

) —

4 (2

.9 )

3 (4

.4 )

— —

In ve

st ig

at io

ns O

rd er

in g

in ve

st ig

at io

ns 39

(6 .2

) 8

(1 8.

6) 6

(3 .4

) 5

(6 .3

) 4

(2 .9

) 8

(1 1.

6) 5

(1 1.

6) 3

(3 .7

) Im

pl em

en ti

ng in

ve st

ig at

io ns

45 (7

.1 )

2 (4

.7 )

9 (5

.1 )

14 (1

7. 5)

13 (9

.4 )

4 (5

.8 )

— 3

(3 .7

) R

ep or

ti ng

in ve

st ig

at io

ns 77

(1 2.

2) 1

(2 .3

) 16

(9 .0

) 4

(5 .0

) 12

(8 .6

) 8

(1 1.

6) 12

(2 7.

9) 24

(2 9.

6) M

ed ic

at io

ns or

O th

er T

re at

m en

ts O

rd er

in g

m ed

ic at

io ns

or tr

ea tm

en ts

a 55

(8 .7

) 8

(1 8.

6) 22

(1 2.

4) —

18 (1

3. 0)

2 (2

.9 )

3 (7

.0 )

2 (2

.5 )

D is

pe ns

in g

m ed

ic at

io ns

or im

pl em

en ti

ng tr

ea tm

en ts

36 (5

.7 )

4 (9

.3 )

14 (7

.9 )

1 (1

.3 )

11 (7

.9 )

6 (8

.7 )

— —

C om

m un

ic at

io n

C om

m un

ic at

io n

w it

h pa

ti en

ts 30

(4 .8

) 1

(2 .3

) 6

(3 .4

) 8

(1 0.

0) 8

(5 .8

) 5

(7 .3

) 2

(4 .7

) —

C om

m un

ic at

io n

w it

h ot

he r

he al

th ca

re pr

ov id

er s

sh ar

in g

pa ti

en t

ca re

a 23

(3 .6

) 3

(7 .0

) 2

(1 .1

) 2

(2 .5

) 10

(7 .2

) 3

(4 .4

) —

3 (3

.7 )

C lin

ic al

K no

w le

dg e

or P

er fo

rm an

ce F

ai lu

re to

fo llo

w st

an da

rd or

re co

m m

en de

d pr

ac ti

ce 12

(1 .9

) 1

(2 .3

) 1

(0 .6

) 2

(2 .5

) 5

(3 .6

) 3

(4 .4

) —

—

a A

cr os

s al

lp ra

ct ic

es ,2

1. 9%

(fi lin

g) ,2

8. 4%

(c ha

rt da

ta ),

40 .0

% (o

rd er

in g

m ed

ic at

io ns

or tr

ea tm

en ts

), an

d 4.

3% (c

om m

un ic

at io

n w

it h

ot he

r he

al th

ca re

pr ov

id er

s sh

ar in

g pa

ti en

t ca

re )

of th

es e

ne ar

m is

se s

w er

e se

co nd

ar ily

at tr

ib ut

ab le

to an

E M

R -r

el at

ed pr

ob le

m .

456 JABFM July–August 2015 Vol. 28 No. 4 http://www.jabfm.org

cycles were used to avert potentially dangerous situations identified by near-miss reports; many ex- pressed surprise about the type and frequency of near-miss errors that occurred in their practice. Indeed, the relatively large volume of near-miss re- ports generated by each practice suggests the impor- tance of developing a systematic approach to process

improvement driven not only by potential for harm but also by frequency of occurrence.

Although our project included a cash bonus to practices for their participation, this did not seem to be an important issue for practices to continue near-miss reporting; the per capita reporting rate did not seem to vary depending on whether the

Table 3. Perceived Severity and Estimated Cost of Selected Near Miss Events in Seven Primary Care Practices

Code # of

Reports Event

Description

Severity Ratinga

Mean (SD)

Likelihood of Adverse Event if Near Miss

not Identifiedb n (%)

Potential Financial

Cost of Event to Patientb

n (%)

Estimated Financial

Cost of Event to Practiceb

n (%)

Five Most Common Near Miss Events 1.1.1 160 Filing problems 51.8 (30.7) 38 (23.8) 12 (7.5) 5 (3.1) 1.1.2 95 Chart data problems 35.4 (29.9) 11 (11.6) 8 (8.4) 4 (4.2) 1.3.2 45 Implementing investigations 52.2 (28.2) 10 (22.2) 4 (8.9) 3 (6.7) 1.3.3 77 Reporting investigations 72.0 (28.3) 38 (49.4) 17 (22.1) 1 (1.3) 1.4.1 55 Ordering medication or treatments 59.1 (29.3) 17 (30.9) 6 (10.9) 5 (9.1) Five Near Miss Events Rated Most Potentially Severe 1.2.1 8 Other specific problems with computer 59.2 (25.2) 5 (62.5) 1 (12.5) 0 (0) 1.3.3 77 Reporting investigations 72.0 (28.3) 38 (49.4) 17 (22.1) 1 (1.30) 1.4.1 55 Ordering medications or treatments 59.1 (29.3) 17(30.9) 6 (10.9) 5 (9.1) 1.4.2 36 Dispensing medications or implementing

treatments 63.0 (31.0) 16 (44.4) 11 (30.6) 3 (8.3)

2.1 12 Failure to follow standard or recommended practice

56.5 (21.7) 2 (16.7) 3 (25.0) 1 (8.3)

Fivec Near Miss Events Rated Most Potentially Costly to the Practice 1.1.4 30 Appointment or referral 46.1 (28.6) 6 (20.0) 4 (13.3) 1.2 8 Equipment and physical building/

surroundings/practice site 49.7 (31.8) 2 (25.0) 1 (12.5)

1.3.2 45 Implementing investigations 52.2 (28.2) 10 (22.2) 4 (8.9) 1.4.1 55 Ordering medication or treatments 59.1 (29.3) 17 (30.9) 6 (10.9) 1.4.2 36 Dispensing medications or implementing

treatments 63.0 (31.0) 2 (44.4) 11 (30.6)

2.1 12 Failure to follow standard or recommended practice

56.5 (21.7) 2 (16.7) 3 (25.0)

a On a scale of 0 –100, where 0�not very serious and 100�extremely serious. b Rated categorically with 0 � none/minimal; 1 � some; 2 � a lot. Percent reflects ‘A lot’ responses within each error type. c Two events tied for fifth most costly. SD, standard deviation.

Table 4. Summary of Practice Improvement Projects Implemented by Participating Primary Care Practices during the Study Period

Primary Care Practice

A B C D E F G

Length of study period in months 8 9 8 9 9 6 6 Number of

Practice Improvement Projects

Initiated 6 6 2 3 15 1 1 Completed in study period 0 0 0 0 7 0 1 Still in process at end of study period 6 0 1 1 4 1 0 On hold at end of study period 0 1 1 1 0 0 0 Inactivated during study period 0 5 0 1 4 0 0

doi: 10.3122/jabfm.2015.04.140050 Using Near-Miss Events to Improve Patient Safety 457

practice offered a reporting incentive. In fact, study practices have continued to log near-miss reports even after the project officially ended and the cash bonuses stopped. Practice leader buy-in and en- couragement seems to be a key element of a suc- cessful reporting system, as has been demonstrated in hospital settings.30

This study has several important limitations. Al- though we purposively chose practices to represent a diversity of size, ownership, specialty, and range of clinical services, our sample was small; therefore results cannot be generalized to all US primary care practices. Similarly, the frequency and types of near-miss reports in this sample cannot be used to estimate the frequency of actual near-miss events. Furthermore, even under the conditions of the study, some underreporting likely occurred. In ad- dition, because event reporting was anonymous, we could not be certain that some events were not reported more than once (ie, by different individ- uals).

Our project involved only near-miss reports. We took great care to exclude AEs (where harm came to the patient) because of concerns of legal liability associated with data sharing. The self-report of likelihood of harm resulting from a near-miss event is, therefore, an estimate. According to leaders in our participating practices, near-miss events that affect patient outcomes are rare; therefore, it is possible either that the subjective estimates were exaggerated or, alternatively, that patients may suf- fer low-level AEs more often than they report. The reporting system did not invite patients to report errors, as some have suggested.31

The reporting phase of our project lasted only 7 to 9 months. The short time frame is insufficient to make broad conclusions about how practice change may result from near-miss reporting or how endur- ing those changes will prove to be. This important question requires further study over a longer period of time.

Conclusions We demonstrated that an anonymous near-miss reporting system can be successfully implemented in a diverse group of primary care practices in a region. The reports generated indicate that near- miss events occur frequently in office practice, pri- marily involve administrative and communication problems, and occasionally pose a significant risk of

patient harm. Practice leaders in our project found these reports helpful and used this information to implement meaningful practice improvement. Fur- ther study is needed to determine whether these improvements can be sustained.

The near-miss reporting and tracking system was developed by Scott Pierson of WindSwept Solutions, Austin, Texas.

References 1. National Ambulatory Med Care Survey: 2010 summary

tables. Available from: http://www.cdc.gov/nchs/data/ ahcd/namcs_summary/2010_namcs_web_tables.pdf. Ac- cessed May 19, 2015.

2. National Learning and Reporting Service, National Patient Safety Agency. Being open: communicating patient safety incidents with patients, their families and carers. Reference number 1097B. November 19, 2009. Available from: http://www.nrls.npsa.nhs.uk/resources/? entryid45�65077. Accessed May 19, 2015.

3. Hoffmann B, Beyer M, Rohe J, Gensichen J, Gerlach FM. “Every error counts”: a web-based incident re- porting and learning system for general practice. Qual Saf Health Care 2008;17:307–12.

4. Bhasale A. The wrong diagnosis: identifying causes of potentially adverse events in general practice using incident monitoring. Fam Pract 1998;15:308 –18.

5. Fernald DH, Pace WD, Harris DM, West DR, Main DS, Westfall JM. Event reporting to a primary care patient safety reporting system: a report from the ASIPS collaborative. Ann Fam Med 2004;2:327–32.

6. Near-miss project e-newsletter. Albany: New York Chap- ter American College of Physicians. Available from: http://www.nyacp.org/i4a/pages/index.cfm?pageid� 3558. Accessed May 19, 2015.

7. Hickner J, Zafar A, Kuo GM, et al. Field test results of a new ambulatory care Medication Error and Adverse Drug Event Reporting System–MEADERS. Ann Fam Med 2010;8:517–25.

8. Kaprielian V, Østbye T, Warburton S, Sangvai D, Michner L. A system to describe and reduce medical errors in primary care. In: Henriksen K, Battles JB, Keyes MA, Grady ML, eds. Advances in patient safety: new directions and alternative approaches. Vol. 1: Assessment. AHRQ publication no. 08 – 0034-1. Rockville, MD: Agency for Healthcare Re- search and Quality; 2008.

9. Zwart DL, Steerneman AH, van Rensen EL, Kalkman CJ, Verheij TJ. Feasibility of centre-based incident reporting in primary healthcare: the SPIEGEL study. BMJ Qual Saf 2011;20:121–7.

10. Elder NC, Graham D, Brandt E, et al. Barriers and motivators for making error reports from family medicine offices: a report from the American Acad- emy of Family Physicians National Research Net- work (AAFP NRN). J Am Board Fam Med 2007;20: 115–23.

458 JABFM July–August 2015 Vol. 28 No. 4 http://www.jabfm.org

http://www.cdc.gov/nchs/data/ahcd/namcs_summary/2010_namcs_web_tables.pdf

http://www.nrls.npsa.nhs.uk/resources/?entryid45=65077

http://www.nyacp.org/i4a/pages/index.cfm?pageid=3558

11. Rowin E, Lucier D, Pauker S, Kumar S, Chen J, Salem D. Does error and adverse event reporting by physicians and nurses differ? Jt Comm J Qual Patient Saf. 2008;34:537– 45.

12. Fisher MA, Mazor KM, Baril J, et al. Learning from mistakes. Factors that influence how students and residents learn from medical errors. J Gen Intern Med 2006;21:419 –23.

13. Chan KD, Gallagher TH, Reznick, et al. How sur- geons disclose medical errors to patients: a study using standardized patients. Surgery 2005;138:851– 8.

14. Hingorani M, Wong T, Vafidis G. Patients’ and doctors attitudes to amount of information given after unintended injury during treatment: cross sectional, questionnaire survey. BMJ 1999;318: 640 –1.

15. Makeham MA, Stromer S, Kidd MR, Lessons from the TAPS study-reducing the risk of patient harm. Aust Fam Physician 2008;37:339 – 40.

16. Pace WD, Staton EW, Higgins GS, Main DS, West DR, Harris DM; ASIPS Collaborative. Database de- sign to ensure anonymous study of medical errors: a report from the ASIPS Collaborative. J Am Med Inform Assoc 2003;10:531– 40.

17. Kennedy AG, Littenberg B, Senders JW. Using nurses and office staff to report prescribing errors in primary care. Int J Qual Health Care 2008;20:238 – 45.

18. Sandars J, Esmail A. The frequency and nature of medical error in primary care: understanding the diversity across studies. Fam Pract 2003;20:231– 6.

19. Rubin G, George A, Chinn DJ, Richardson C. Er- rors in general practice: development of an error classification and pilot study of a method for detect- ing errs. Qual Saf Health Care 2003;12:443–7.

20. Kennedy AG, Littenberg B, Senders JW. Using nurses and office staff to report prescribing errors in primary care. Int J Qual Health Care 2008;20:238 – 45.

21. West DR, Pace WD, Dickinson LM, et al. Relation- ship between patient harm and reported medical errors in primary care: a report from the ASIPS Collaborative. In: Henriksen K, Battles JB, Keyes MA, Grady ML (eds). Advances in patient safety: new directions and alternative approaches. Vol. 1:

Assessment. Rockville, MD: Agency for Healthcare Research and Quality; 2008.

22. Elder NC, Vonder Meulen M, Cassedy A. The identification of medical errors by family physi- cians during outpatient visits. Ann Fam Med 2004; 2:125–9.

23. Woods DM, Thomas EJ, Holl JL, Weiss KB, Bren- nan TA. Ambulatory care adverse events and pre- ventable adverse events leading to a hospital admis- sion. Qual Saf Health Care 2007;16:127–31.

24. Aspden P, Corrigan JM, Wolcott J, et al. Patient safety: achieving a new standard for care. Washing- ton, DC: National Academies Press; 2004.

25. Elder NC, Vonder Meulen M, Cassedy A. The iden- tification of medical errors by family physicians dur- ing outpatient visits. Ann Fam Med 2004;2:125–9.

26. Dovey SM, Meyers DS, Phillips RL Jr, et al. A preliminary taxonomy of medical errors in family practice. Qual Saf Health Care 2002;11:233– 8.

27. Bates DW, Cohen M, Leape LL, Overhage JM, Shabot MM, Sheridan T. Reducing the frequency of errors in medicine using information technology. J Am Med Inform Assoc 2001;8:299 –308.

28. Harrington L, Kennerly D, Johnson C. Safety issues related to the electronic medical record (EMR): syn- thesis of the literature from the last decade, 2000 – 2009. J Healthc Manag 2011;56:31– 44.

29. Nanji KC, Rothschild JM, Salzberg C, et al. Errors associated with outpatient computerized prescribing system. J Am Med Inform Assoc 2011;18:767–73.

31. Simmons D, Mick J, Graves K, Martin S. 26,000 Close call reports: lessons from the University of Texas Close Call Reporting System. In: Henriksen K, Battles JB, Keyes MA, Grady ML (eds). Advances in patient safety: new directions and alternative approaches. Vol. 1: As- sessment. Rockville, MD: Agency for Healthcare Re- search and Quality; 2008.

32. Phillips RL, Dovey SM, Graham D, Elder NC, Hickner JM. Learning from different lenses: reports of medical errors in primary care by clinicians, staff, and patients: a Project of the American Academy of Family Physicians National Research Network. J Pa- tient Saf 2006;2:140 – 6.

doi: 10.3122/jabfm.2015.04.140050 Using Near-Miss Events to Improve Patient Safety 459

Appendix Appendix A: Near-Miss Reporting Form [Adapted for single-page display]

Near-Miss Reporting Form for Physicians and Staff Date: __ __ / __ __ / __ __ All Near-miss Reporting forms are completely Anonymous and Confidential mo day year

Use this form for NEAR-MISS reporting only if no significant harm came to the patient as a result of the event. Actual injury needs to be reported via other channels.

1. Is the event related to a specific patient? Yes If yes � please answer Questions 2-5 No If no � please skip to Question 6

2. What are your past experiences with this patient (select one response)? NEVER seen pt + NOT familiar w health problems HAVE seen pt but NOT familiar w health problems SOMEWHAT familiar with the pt and health problems QUITE familiar with the pt and health problems VERY familiar with the pt and health problems

3. What is the patient's age (estimate if unsure; if <2 years, state age in months). ____ ____ ____ years months

4. Does this patient have a CHRONIC health problem (select one response)? Yes No Don't know

Comment on this question:

5. Does this patient have a COMPLEX health problem (select one response)? Yes No Don't know

Comment on this question:

6. What happened? (Please think about what, where, and who was involved. DO NOT USE NAMES OR DATES IN YOUR ANSWERS).