G Model YDLD-2200; No. of Pages 5

ARTICLE IN PRESS Digestive and Liver Disease xxx (2012) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld

Digestive Endoscopy

Errors and near misses in digestive endoscopy units Giorgio Minoli a,∗ , Paolo Borsato b , Enrico Colombo c , Aurora Bortoli d , Tino Casetti e , Giovanni de Pretis f , Luca Ferraris g , Ivano Lorenzini h , Alberto Meggio i , Rudy Meroni j , Lucia Piazzi k , Vittorio Terruzzi a a

UO di Gastroenterologia, Ospedale Valduce, Como, Italy Dipartimento di Studi Sociali e Politici, Università di Milano, Italy UO di Gastroenterologia, AO Salvini, Ospedale di Garbagnate Milanese, Milan, Italy d UO di Gastroenterologia, AO Salvini, Ospedale Rho, Milan, Italy e UO di Gastroenterologia, Ospedale di Ravenna, Italy f UO di Gastroenterologia, Ospedale Santa Chiara, Trento, Italy g UO di Gastroenterologia, Ospedale S. Antonio Abate, Gallarate, Varese, Italy h UO di Gastroenterologia, Ospedali Riuniti, Ancona, Italy i UO di Gastroenterologia, Ospedale SM del Carmine, Rovereto, Trento, Italy j Centro Elaborazione Dati, Ospedale Valduce, Como, Italy k UO di Gastroenterologia, Ospedale Centrale di Bolzano, Italy b c

a r t i c l e

i n f o

Article history: Received 13 March 2012 Accepted 17 June 2012 Available online xxx Keywords: Endoscopic workflow Errors Incident reporting Near misses

a b s t r a c t Background: Not much is known about errors and near misses in digestive endoscopy. Aims: To verify whether an incident report, with certain facilitating features, gives useful information about unintended events, only excluding errors in medical diagnosis. Method: Nine endoscopy units took part in this cross sectional, prospective, multicentre study which lasted for two weeks. Members of the staff were required to report any unintended, potentially dangerous event observed during the daily work. A form was provided with a list of “reminders” and facilitators were appointed to help. The main outcome measurements were type of event, causes, corrective interventions, stage of occurrence in the workflow and qualification of the reporters. Results: A total of 232 errors were reported (two were not related to endoscopy). The remaining 230 amount to 10.3% of 2239 procedures; 66 (29%) were considered errors with consequences, 164 (71%) “near misses”. There were 150 pre-operative errors (65%), 22 operative (10%) and 58 post-operative (25%). Corrective interventions were provided for 60 cases of errors and 119 near misses. Most of the events were reported by the nurses (106 out of 232, 46%). Conclusions: Short-term incident reporting focusing on near misses, using forms with lists of “reminders”, and the help of a facilitator, can give useful information on errors and near misses in digestive endoscopy. © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

1. Introduction From 3% to 16% of patients suffer harm during treatment in hospital; half these episodes are avoidable, and 13% are fatal [1]. Increasing efforts to avoid errors have been made in recent years but there is still room for improvement [2,3]. Quality assurance is a very import intervention to prevent damages and dissatisfaction of the patients and one of the first steps is to recognize and report the errors in order to allow their prevention (audit, check list and so on).

∗ Corresponding author at: Via Molino 20, 22030 Montorfano (CO), Italy. Tel.: +39 031 200474; fax: +39 031 200474. E-mail address: [email protected] (G. Minoli).

To our knowledge, no investigation specifically designed to report errors and near misses in digestive endoscopy has been published so far. We used data based on incident reporting to study all possible errors, only excluding errors in medical diagnosis. We also studied the distribution of the errors in the different phases of the workflow (pre-, during or post-procedure) to identify the phase requiring most attention [4,5]. One of the difficulties of using incident reporting is its weakness in gathering all errors [6], quite likely because of fear of punishment. To overcome these obstacles we developed a less “guilt-provoking” incident reporting procedure which allowed for reports of errors but focused more on near misses, which people are less likely to hesitate to report, but which are still important if the goal is to improve safety [7,8]. We made it short, to be as convenient as possible, and a “facilitator” was always present to help, with a “reminder”

1590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dld.2012.06.009

Please cite this article in press as: Minoli G, et al. Errors and near misses in digestive endoscopy units. Dig Liver Dis (2012), http://dx.doi.org/10.1016/j.dld.2012.06.009

G Model YDLD-2200; No. of Pages 5

ARTICLE IN PRESS G. Minoli et al. / Digestive and Liver Disease xxx (2012) xxx–xxx

2

form listing the most likely events to jog the compilers’ memories; absolute anonymity was guaranteed. The aim of the study was to further our knowledge of critical events in endoscopy not depending on the physician’s technical performance of the endoscopic examination, using an incident reporting form with several facilitating features.

2. Materials and methods 2.1. Participating units We made no preliminary selection of the participating endoscopy units. Simply, on various occasions (for instance at the meetings of the scientific societies), we communicated our intention to conduct the survey and accepted requests for participation. The size of the units in 2008 was deduced by the number of examinations, endoscopists and nurses and by the presence of administrative personnel and of a Central Booking Service.

They are listed in Table 1.

2.3. Organization On 19 September 2009 a meeting of the department heads was held and the project was discussed. Special attention was paid to the definition of “near misses” and to drafting the form with the reminders listed as “memory joggers”. Before starting the survey a facilitator was nominated in each unit, and copies of the form were then distributed throughout the premises of the single units (endoscopy rooms, instrumentation cleaning rooms, areas for discharging patients, administrative offices, and so on). Formal written authorization from ethics committees was requested and obtained only at Gallarate and Valduce hospitals; the other committees felt it was unnecessary because patients were not directly involved.

Table 1 Definitions.

Near miss

Error

Workload Pre-operative error Operative error

Post-operative error

Recording lasted two weeks, to be selected by each unit between 15 November and 15 March, with the exception of the period from 15 December to 15 January (Christmas and New Year holidays). Each staff member (endoscopist, nurse, nursing auxiliary, secretary, and administrative personnel) was required to enter on the form any unintended event, with its causes, corrective measures, and consequences, they encountered during their usual working time, whether listed on the form as a reminder or not. Anonymity was guaranteed. The “facilitator” (M.D. and/or nurse) had to remind everyone in the team that they were expected to record any events they noticed. Recording and describing the events depended exclusively on the observer’s own judgement. A second judgement by a group of experts would have made the study more difficult and dissimilar from the usual way of working. This meant, however, that there was no way of avoiding the same event being considered differently as worthy of mention and as an error or near miss, not only in the different units but even within the same one. 2.5. The form

2.2. Definitions

Event

2.4. Data recording

Any unintended event in the reporting staff member’s opinion (for example: patient not fasted). An event classified by those reporting it as: (1) potentially but not actually happening; (2) potential but promptly intercepted; (3) no causing any harm to the patient. An event that caused: (4) minor harm (postponement, significant delay); (5) moderate harm (requiring further medical investigation or pharmacological treatment); (6) moderate-significant harm (involving major treatment and/or adjustment to the working programme); (7) significant harm (hospital admission or a longer stay); (8) severe injury (permanent disability, contribution to death). The number of gastroscopies and colonoscopies done in the two weeks of the survey. Any event occurring between the patient being admitted and positioning on the endoscopy bed. Any event occurring between the patient being positioned on the endoscopy bed and leaving the endoscopy room. Any event occurring between the patient leaving the endoscopy room and discharge from the endoscopy unit.

The form was divided into six sections; four listed items that might need to be reported (reminders) and two did not, but called for details of the event(s) and any corrective measures adopted. The reminders list included: the most likely events on which to focus (Table 3), the consequences (in order to distinguish near misses and mistakes), the causes of mistakes (Table 4) and the member of the staff who reported the events. 2.6. Classification of events in relation to the workflow At the end of the survey we classified the events reported on the forms in relation to the stage in the workflow as: pre-operative, operative and post-operative (see Section 2.2). 3. Results 3.1. Participating units (Table 2) Twelve units asked about the survey, and nine eventually participated. They were at the following hospitals: Ancona, Bolzano, Valduce (Como), Gallarate (Varese), Garbagnate Milanese (Milan), Ravenna, Rho (Milan), Rovereto (Trento) and Trento. Overall, 53,200 endoscopic examinations (gastroscopies and colonoscopies) were performed in year 2008. The mean number Table 2 Workload and number of events reported by the single units. Unit A B C D E F G H I Total

Workload 311 150 287 318 270 190 231 222 260 2239

Events no. (%)a

Errors no. (%)b

Near misses no. (%)c

39 (13) 8 (5) 58 (20) 36 (11) 29 (11) 8 (4) 4 (2) 19 (9) 31(12)

25 (64) 5 (63) 11 (19) 2 (6) 5 (17) 3 (38) 3 (75) 4 (21) 8 (26)

14 (36) 3 (37) 47 (81) 34 (94) 24 (83) 5 (62) 1 (25) 15 (79) 23 (74)

232 (10.4)

66 (28)

166 (72)

The order of the units is not the same as in the other tables, to maintain anonymity; events not related to endoscopy are included. a Number of events, and percentage of the workload. b Number of errors, and percentage of the number of events. c Number of near misses, and percentage of the number of events.

Please cite this article in press as: Minoli G, et al. Errors and near misses in digestive endoscopy units. Dig Liver Dis (2012), http://dx.doi.org/10.1016/j.dld.2012.06.009

G Model YDLD-2200; No. of Pages 5

ARTICLE IN PRESS G. Minoli et al. / Digestive and Liver Disease xxx (2012) xxx–xxx

of procedures per centre was 5911 (SD ± 990, range 4600–7600). Seven centres had dedicated personnel and eight a central booking service. The mean number of endoscopists and nurses per centre was 6.2 (SD ± 3.1, range 2–13) and 2.9 (SD ± 2.9, range 4–12), respectively. 3.2. Types and numbers of events reported (Table 3) Despite the emphasis on reporting near misses, errors were reported as well. Overall 232 forms were completed and 232 events reported; two of them (failure to close a fridge door and wrong urea breath test result) were not related to endoscopy. The remaining 230 accounted for 10.3% of 2239 procedures. 66 (29%) were considered mistakes (with consequences), 164 (71%) “near misses”. Six did not appear in the reminders list. The numbers of events were not related to the workload of the single units and ranged from 2% to 20%. Doctors reported 65 events (28%), 25 errors and 40 near misses; nurses 106 (46%), 30 of them errors; auxiliary personnel six (3%),

3

all near misses; administrative staff reported 55 events (24%), 11 errors and 44 near misses. 3.3. When mistakes occurred In relation to the stage in the workflow the events were classified at the end of the survey under the following headings (see Section 2.2): 150 (65%) pre-operative (39 errors and 111 near misses), 22 (11%) operative (9 errors and 13 near misses) and 58 (25%) postoperative (14 errors and 44 near misses). 3.4. Causes of errors (Table 4) As more than one cause could be entered for each event, the total exceeded the number of errors (360 entries for 232 errors); they referred to 22 of the 26 items listed as reminders on the survey form. Individual oversight was entered 92 times, failure to follow the procedure 83 times, and little importance attributed to the mistake 61 times. Other possible causes were entered less often.

Table 3 Reminders, reported events and their position in the workflow. Reminders and events Pre-operative Lack of informed consent form or mistakes in its compilation (at the beginning of the endoscopy) Failure to give the patient the information form about the procedure (in accordance with current procedures) Loss of or incomplete general practitioner’s (GP) prescription form Errors in the booking date/time of the exam Errors on labels (wrong name, surname, date of birth) made in the administrative section No biochemistry tests for blood coagulation (in accordance with the endoscopy unit’s current procedures, e.g. for a PEG) Patient allocated to an investigation different from that requested by the GP Patient not fasted Patient’s dental prostheses not removed No warning of anticoagulant treatment No warning of clinical setting which could interfere with drugs for sedation/analgesia Wrong patient on the endoscopy bed Intravenous sedation prepared wrongly Patient has not taken the preparation for colonoscopy Mistaken identity of the patient No warning of treatment with aspirin or anti-inflammatory drugs Total of pre-operative errors Operative Malfunction of the endoscopic instrumentation because of lack of pre-procedure testing Missed or wrong recording of the drugs administered Opening the wrong attachment/accessory No sample for histology (e.g. failure to keep a polyp that was removed) Use of non-sterile instrument (in accordance with the unit’s current procedure) Lack of the accessories needed to complete the procedure (pliers, loops, needles, etc.) Wrong drug administered (type and/or dosage) for sedation/analgesia Unintentional use of the wrong kind of electric current Total of operative errors Post-operative d Incomplete nursing form Incomplete endoscopic report: wrong administrative data Request for pathology examination compiled wrongly Incomplete endoscopic report: no mention of any biopsies or similar that were done Incomplete endoscopic report: procedure done not correctly described Incomplete endoscopic report: missed or wrong mention of the drug administered (type or dose) Incomplete report: lack of information for correct management of the patient after the endoscopy Errors in labelling the samples for histology No report of the patient’s personal documents or belongings Failure to return personal belongings to the patient (or accompanying persons) Errors in delivering the endoscopic report (to a wrong or unauthorized person) d Mismanagement of the endoscopic report d Loss of the patient’s documentation Failure to return personal health papers to the patient or accompanying persons d Wrong management of the histological report Total post-operative errors Total endoscopy-related events

E no. (%)a

Er no. (%)b

NM no. (%)c

39 (17) 30 (13) 22 (9) 17 (7) 7 (3) 5 (2) 5 (2) 5 (2) 4 (2) 3 (1) 3 (1) 3 (1) 2 (1) 2 (1) 2 (1) 1 (0.5) 150 (65)

– – – 12 (71) 4 (57) 5 (100) 3 (60) 5 (100) – 3 (100) 1 (33) 1 (33) 2 (100) 2 (100) 1 (50) – 39 (26)

39 (100) 30 (100) 22 (100) 5 (29) 3 (43) – 2 (40) – 4 (100) – 2 (67) 2 (67) – – 1 (50) 1 (100) 111 (74)

4 (2) 4 (2) 4 (2) 3 (1) 3 (1) 2 (1) 1 (0.5) 1 (0.5) 22 (10)

2 (100) 3 (75) 2 (50) 3 (100) 2 (67) 1 (50) – – 13 (59)

2 (100) 1 (25) 2 (50) – 1 (33) 1 (50) 1 (100) 1 (100) 9 (41)

13 (6) 11 (5) 5 (2) 4 (2) 3 (1) 3 (1) 3 (1) 3 (1) 3 (1) 2 (1) 2 (1) 2 (1) 2 (1) 1 (0.5) 1 (0.5) 58 (25) 230

– 7 (64) – 2 (100) – 1 (33) 2 (67) – 1 (33) – 1 (50)

13 (100) 4 (36) 5 (100) 2 (100) 3 (100) 2 (67) 1 (33) 3 (100) 2 (67) 2 (100) 1 (50) 2 (100) 2 (100) 1 (100) 1 (100) 44 (76) 164 (71)

– – – 14 (24) 66 (29)

E: events; Er: errors; NM: near misses. a Number of single events as a percentage of the total reported events. b Number of errors as a percentage of the reported events. c Number of near misses as a percentage of the reported events. d Events not listed among the reminders.

Please cite this article in press as: Minoli G, et al. Errors and near misses in digestive endoscopy units. Dig Liver Dis (2012), http://dx.doi.org/10.1016/j.dld.2012.06.009

G Model YDLD-2200; No. of Pages 5

ARTICLE IN PRESS G. Minoli et al. / Digestive and Liver Disease xxx (2012) xxx–xxx

4 Table 4 Causes in the observer’s or facilitator’s opinion. Cause Individual “oversight” Rule/procedure not followed Little importance attributed to such mistakes Chaotic activity Colleagues from other wards distracted Collective “oversight” Patient distracted Scant communication No rule/procedure available Inadequate physical or technological setting Underestimation of risk Stress Scant mutual help Disorganized flow of patients Inadequate supervision Relative distracted No other possible solution Incomplete training Pressure for high productivity Need to comply with the daily work schedule Total

Frequency no. (%)a 92 (26) 83 (23) 61 (17) 21 (6) 20 (5) 16 (4) 14 (4) 13 (4) 12 (3) 6 (2) 5 (1) 3 (1) 3 (1) 3 (1) 2 (0.6) 2 (0.6) 1 (0.2) 1 (0.2) 1 (0.2) 1 (0.2) 360 (100)

ER no. (%)b

NM no. (%)c

30 (33) 9 (11) 5 (8) 10 (48) 8 (40) 10 (63) 6 (43) 6 (46) 3 (25) 1 (17) – 1 (33) 1 (33) 3 (100) 1 (50) 2 (100) 1 (100) – – –

62 (67) 74 (89) 56 (92) 11 (52) 12 (60) 6 (37) 8 (57) 7 (54) 9 75) 5 (83) 5 (100) 2 (67) 2 (67) – 1 (50) – – 1 (100) 1 (100) 1 (100)

97 (27)

263 (73)

M: mistakes; NM: near misses. a Number and percentage of all the causes. b Number and percentage of causes of errors. c Number and percentage of causes of near misses.

Stress was mentioned three times; tiredness, de-motivation, difficult decision, need to reduce discomfort of waiting patients, help colleagues in difficulty and make room for an emergency – which also were offered in the reminders – were never reported.

3.5. Corrective measures (Table 5) Corrective measures were taken for 60 errors and 119 near misses; no corrective measures were indicated for 47 events, and for four errors and two near misses they did not seem appropriate.

The number and severity of the events reported were sometimes limited and had no important consequences but that does not mean they were trivial, because their potential for harm and the need for prevention must never be neglected. Incident reporting, at least on the basis of published experience, is far from satisfactory, as not all the negative events that arise are reported [6,10]. These weak points were confirmed, for example, by Kaldjian et al. [11] who, using a questionnaire, reported that 84%

Table 5 Corrective measures. No. (%)

4. Discussion Our aim was to study if an incident reporting performed for a short time, with the help of a form (with reminders of the events to be reported) and of a facilitator could give useful information on errors and near misses, on their causes and corrective measures. We also intended to record the stage of errors in the workflow. We confirmed that many errors can happen in digestive endoscopy units, and that even short-term incident reporting can give useful information. In all, 232 events were reported, 230 really concerning endoscopy, corresponding to 10.3% of the workload, and the majority were near misses (72%). The units differed widely in size and staff, so the findings can probably apply to a wide range of endoscopy units, at least in Italy. The numbers and type of events reported also differed widely from one unit to another (Table 2), with no relation to their size (indicated by the workload). This finding is discouraging, highlighting differences in attitudes to dealing with preventing mistakes. The most frequent events were “process errors” [9] mostly involving contact with patients (informed consent, information on the procedure, mistakes in the booking date or time); technical ones (no blood coagulation tests, malfunction of the endoscopic instrumentation, or the use of non-sterile instruments) were far less frequent, probably meaning that the staff tend to concentrate mainly on the examination itself and not so much on other aspects of the endoscopic work, forgetting that attention must be paid to the whole cycle.

In case of errors Re-scheduling the examination on another date or at another time Administrative interventions (correction of administrative documentation and booking records, contacts with patients at home) Correction of the GP’s request Corrections of the endoscopic reports Replacement of inappropriate accessories New preparation for sedation (different drugs, insufficient information about the clinical situation of the patients) New contact with the pathological department to correct the histological request Repetition of the examination Contacting the patient to return documents Total measures for correcting errors In case of near misses Correction of the informed consent form Repetition of the information procedure Completion of the nursing form Correction of the endoscopic report Correction of the GP’s request Replacement of wrong endoscopy accessories laid out before the examination Re-scheduling the examination Correction of mistakes in the histological request and new contact with the pathological department New administrative action Contacting patients to return documents Correction of sedation preparation Total measures for correcting near misses

18 (30) 15 (25)

8 (13) 4 (7) 4 (7) 4 (7) 4 (7) 2 (3) 1 (2) 60 (100) 32 (27) 30 (25) 12 (10) 11 (9) 8 (7) 7 (6) 6 (5) 4 (3) 4 (3) 4 (3) 1 (1) 119 (100)

Please cite this article in press as: Minoli G, et al. Errors and near misses in digestive endoscopy units. Dig Liver Dis (2012), http://dx.doi.org/10.1016/j.dld.2012.06.009

G Model YDLD-2200; No. of Pages 5

ARTICLE IN PRESS G. Minoli et al. / Digestive and Liver Disease xxx (2012) xxx–xxx

of 388 doctors agreed that “reporting improves quality” but only 17.8% did actually report a minor mistake and 3.8% a major one. Despite these weaknesses, however, it remains a useful means of gathering extra information, in addition to what other systems provide. It might be strengthened, with limited effort and training, by encouraging workers to report important critical events. If a light hand is used it can be repeated – even several times – so as to make up to some extent for the lack of completeness. We felt the focus on near misses was an important facilitating feature, as these are closely related to true mistakes. For instance, in a study on bladder tumours 65 near misses were reported out of 115 events (56.5%) and their collection permitted easier analysis and prevention [12,13]. The use of the form with reminders, intended to jog the compilers into remembering things, may, however, have influenced the choice of events to be reported and this is a limitation. The survey indicated that, in an endoscopy unit, process errors and near misses fell mainly into the pre-operative phases, most in the very early contacts with the patient. The first four on our list – lack of informed consent; failure to give the patient information about the procedure; incomplete GP prescription; errors in the booking date/time of the examination – explained nearly half the events reported – 46%. Errors and near misses in these stages are serious because they can easily lead to problems in later phases, and in any case interfere with work planning. Unfortunately the design of our trial was not specifically addressed to investigate for measures to avoid errors and near misses, as for example a pre-examination discussion between patients and endoscopists or nurses. There were fewer operative events, concerning malfunctioning of instruments or wrong choice of accessories and mistakes in sedation (five cases). Post-procedural errors [14] were more frequent and involved the physicians’ or nurses’ reports and, again, communication with patients or their relatives. The types of event and their distribution in the workflow can suggest the phase needing the closest attention (in each unit) when planning prevention schemes. On the question of the causes of errors, it is interesting that those involving individuals were far fewer than the collective and organizational ones. There were 92 reports of individual oversight, but only 51 were considered the sole cause of error – less than a quarter of the events reported. Stress, fatigue and lack of motivation scored very low. Most events were considered to involve more than one cause, comprising attitudes – such as little importance attributed to the type of error – or organizational problems such as chaotic work, lack of communication, distraction by colleagues in other wards, etc. Almost half the reports – 155 out of 360 – fell under these two headings. Failure to respect the rules (83 reports) can be interpreted as organizational slackness. These indications confirm the organizational setting as a key area where risk prevention might have most effect. The findings also suggest that, at least for near misses, unwillingness to report the events because of fear of punishment might not always really be the main explanation. The corrective measures as a whole seem to imply an impact on regular workflow, causing organizational hitches. Correcting errors, re-scheduling or changing the sequence of work (from postponing the examination to doing non-scheduled tasks) was frequently necessary. For the near misses the corrective measures, although often only small, almost always had an impact on the continuity of the workflow, disrupting “downstream” activities.

5

Possibly because our survey was so short, we have no reports of analysis or changes intended to improve the procedures and prevent the recurrence of mistakes. Herein lies a further utility of incident reporting, short if needed, but repeated: it can highlight frequent risks that call for attention and action. The medical staff showed limited interest, reporting only 65 events, while non-medical personnel seemed more active and reported many more. This too suggests reflection on which categories to concentrate most effort. In conclusion, this survey, limited to endoscopy units, found numerous errors and near misses, most of them in the preoperative stages. They affect the regular workflow and seem mainly to be caused by “systemic” organizational problems. The findings suggest that more focus is needed on the process flow and on the structural context, in agreement with the systemic view of mistake prevention proposed in the literature. Conflict of interest statement None declared. Collaborators Simona Balzani (Ravenna), Elena Calvetti (Rovereto, TN), Donato Casagrande (Garbagnate, MI), Fausto Chilovi (Bolzano), Giuseppe Costa (Como), Cristiano Crosta (Milano), Katia Faitini (Trento), Giancarla Fiori (Milano), Sara Ierace (Bolzano), Gianni Imperiali (Como), Mauro Medas (Como), Luisa Micheletto (Gallarate, VA), Gianni Miori (Trento), Alberto Prada (Rho, MI), Lorenza Sartirana (Rho, MI), Laura Beatriz Siena (Ancona). References [1] Institute of Medicine. To err is human. Washington, DC: National Academy Press; 1999. [2] Leape L, Berwick DM. Five years after to err is human: What have we learned? Journal of the American Medical Association 2005;293:2384–90. [3] Vincent C, Aylin P, Franklin BD, et al. Is healthcare getting safer? British Medical Journal 2008;337:a2426, http://dx.doi.org/10.1136/bmj.a2426. [4] Brideau LP. Flow: why does it matter? Frontiers of Health Services Management 2004;20:47–50. [5] Saunders C, Dempsey C. Improving patient flow. Physician involvement drives success at Georgia hospital. Healthcare Executive 2008;23:46–8. [6] Vincent C. Incident reporting and patient safety. British Medical Journal 2007;13:334–53. [7] Barach P, Small SD. How the NHS can improve safety and learning. By learning free lessons from near misses. British Medical Journal 2000;320:1683–4. [8] Barach P, Stephen D, Small SD. Reporting and preventing medical mishaps: lessons from non medical near miss reporting systems. British Medical Journal 2000;320:759–63. [9] Dovey SM, Meyers DS, Phillips Jr RL, et al. A preliminary taxonomy of medical errors in family practice. Quality & Safety in Health Care 2002;11:233–8. [10] Burkoski V. Identifying risk: the limitations of incident reporting. Canadian Nurse 2007;103:12–4. [11] Kaldjian LC, Jones EW, Wu BJ, et al. Reporting medical errors to improve patient safety: a survey of physicians in teaching hospitals. Archives of Internal Medicine 2008;14:40–6. [12] Singh R, Saleemi A, Walsh K, et al. Near misses in bladder cancer—an airline safety approach to urology. Annals of the Royal College of Surgeons of England 2003;85:378–81. [13] Lamb BW, Nagpal K. Patient safety. Importance of near misses. British Medical Journal 2009;339:b3032, http://dx.doi.org/10.1136/bmj.b3032. [14] Nagpal K, Poyhj Arora S, Abboudi M, et al. Postoperative handover: problems, pitfalls, and prevention of error. Annals of Surgery 2010;252:171–6.

Please cite this article in press as: Minoli G, et al. Errors and near misses in digestive endoscopy units. Dig Liver Dis (2012), http://dx.doi.org/10.1016/j.dld.2012.06.009