Improving care safety by characterizing task interruptions during interactions between healthcare professionals: an observational study

Abstract

Few studies have investigated interruptions to the work of professionals practicing in inpatient hospitals, and even fewer take account of the functions that make up the system. Safety of care can be improved by considering avoidable interruptions during interactions between managerial and care delivery functions. The present study describes the characteristics of interruptions to the work of professionals working in the inpatient hospital sector, with respect to their typology, frequency, duration, and avoidability in the context of interactions between functions. This direct observational study of interruptions in hospital care was performed in the Pays de la Loire (west coast) area of France. A total of 23 teams (17 institutions) working in medical or surgical specialties (excluding intensive care) were included. Observations were performed between May and September 2019, and lasted seven consecutive hours per team. A pair of observers simultaneously observed the same professional for ∼30 min. Each occupational category was examined. Reported characteristics were: (i) the method and duration of the request, (ii) the location of interrupted and interrupting persons, (iii) the reaction of the interrupted person, (iv) the characteristics of the interrupting person, and (v) the classification of interrupted and interrupting tasks according to their function. An avoidable interruption was defined. Interruptions during interactions between professionals were categorised in terms of their function and avoidability. Descriptive statistical analyses (mean, standard deviation, and distribution) were run. In particular, cross-comparisons were run to highlight avoidability interruptions and interactions between managerial and care delivery functions during the working day, for different professional categories, and for the location of the request. Overall, 286 interrupted professionals were observed and 1929 interruptions were characterised. The majority of interruptions were due to a face-to-face request (58.7%), lasting ≤30 s (72.5%). Professionals engaged in the response in 49.3% of cases. A total of 57.4% of interruptions were avoidable. The average number of interruptions was 10.5 (SD = 3.2) per hour per professional. An analysis of avoidability and interactions between managerial and care delivery functions found that the period between 12:00 and 13:00 was the riskiest in terms of care safety. This study highlighted the characteristics of interruptions to the activity of professionals working in inpatient hospitals. Care teams could focus on making medical and nursing professionals much more aware of the importance of interruptions, and each team could decide how to best-manage interruptions, in the context of their specific working environment.

Introduction

Although a task interruption (denoted here as an ‘interruption’) is necessary in the context of alerting colleagues to problems and interactions within care processes [1], it is also described in the literature as detrimental [1]. Several articles describe an increased mental load, impaired decision-making, the delayed or forgotten transmission of information, and the increased risk of human error [2, 3]. Other studies have sought to analyse errors and clinical consequences of interruptions, based on a hypothesis of a risk to patient safety [1, 4–6]. Hence, the consideration of interruptions that affect the work of professionals in healthcare institutions is an element of care risk management [1, 4–6]. Although the international literature on interruptions is both vast and multidisciplinary [7], there are several limitations. The first is the lack of a consensus on the concept and the definition of an interruption [4, 7, 8]. The second is the lack of research, and a clear definition of the avoidability of interruptions. To date, healthcare safety research has only proposed definitions of avoidability with reference to Adverse Events Associated with Health Care [9]. A third limitation is the compartmentalised nature of studies, which often target a single professional category [2, 6, 10–12] or specific activities [5, 13–16]. While a mono-disciplinary approach can be instructive, it does not take into account the interdisciplinary dependencies that make up the healthcare system [17]. Furthermore, few studies have investigated interruptions among medical or surgical teams (excluding intensive care) in the inpatient hospital sector [18].

The observational tool (Team’IT), derived from the Dual Perspectives Method proposed by McCurdie et al. [19] makes it possible to examine how different teams function in the inpatient hospital sector [19, 20]. Interruptions are examined from the point of view of the functions that make up the system. ‘Functions’ are defined as what must occur for a system to fulfil its care objectives. They are performed by professionals (independent of their role) and by intelligent agents [19], and are executed through a set of interconnected tasks. An interruption occurs when a task being carried out by the interrupting professional interferes with a task being carried out by the interrupted professional [19]. Tasks are categorised into four functions, defined by Miller et al. [20, 21]: unit resource coordination, care coordination, patient care planning, and patient care delivery. The patient care delivery function is the final barrier; it is closest to the patient and it must receive special attention to ensure patient safety [20]. The remaining functions (coordination and planning) must be anticipated upstream of care delivery, to ensure that patient safety is not the sole responsibility of so-called last-line actors. In practice, healthcare safety relies on the consideration of interactions between managerial (coordination and planning) functions and patient care delivery [20]. Together, these elements also contribute to the question of whether interruptions can be avoided.

The main aim of this study was to characterise interruptions in the inpatient hospital sector with respect to their typology, frequency, duration, and avoidability in the context of interactions between managerial and care delivery functions. The first step was to define avoidable interruptions.

Method

Design

This was a direct observational study of interruptions during the day-to-day activities of healthcare professionals.

Setting

The study was run in the Pays de la Loire region (west coast) of France. A total of 23 teams (at 17 institutions) working in inpatient hospital care participated. Inclusion took place in November 2018, and teams were characterised by their specialty (general medicine or surgery, excluding intensive care), patient length of stay (longer than one day) and accommodation (at least one night) [22]. Inclusion criteria and the characteristics of teams are available elsewhere [20].

Definitions

The definition of an interruption is that used by French health authorities (the Haute Autorité de Santé): ‘the unexpected, temporary or definitive cessation of a human activity. The reason can either be specific to the operator, or, on the contrary, be external. The interruption disrupts the course of the activity, […]. The potential need to carry out secondary activities impedes the completion of the initial activity’ [18].

Both the French [9] and the international [23] literature were consulted to establish the definition of an ‘avoidable interruption’. Here, the aim was to adapt the definition of the term preventable (‘accepted by the community as avoidable in the particular set of circumstances’) provided by the World Health Organization [23] to the context of interruptions occurring in healthcare facilities. The goal was to focus on the quality and safety of care, seen as an organisation in which zero errors would be made. An avoidable interruption was defined in May 2019 as follows: ‘The request or the task that causes the interruption is not necessary for patient safety and could occur at another time. Avoidance of the occurrence of the interruption is achievable through a modification to at least one professional practice (technical and non-technical skills) or the modification of an organizational axes’ (Table 1). Organisational axes were defined as human factors, technical factors, the care environment, and care delivery.

Table 1.

Definition of the four levels of avoidability of task interruption in the context of interactions.

Character of avoidability	Definition	Categorisation of the interrupted task according to function	Examples
Avoidable	The request or the task that causes the interruption is not necessary for patient safety and could be performed at another time. Avoiding the occurrence of the interruption is achievable by changing at least one professional practice (technical and non-technical skills) or changing an organisational axis.	Patient care delivery	Untargeted or poorly targeted request for information; unplanned material needed for planned care; (internal request for information addressed to the wrong person); late or incomplete delivery of important patient information.
		Patient care planning	Non-urgent request for information on the organisation of care within the team; search for a document (a prescription); handover of non-urgent information about other patients.
		Care coordination	Interruptions during shift handovers (by telephone—suppliers—unconcerned professionals) except in an emergency.
		Unit resource coordination	Non-urgent communication of administrative information (e.g. filing documents).
Probably avoidable	The request or the task that causes the interruption is related to patient safety, but could probably be taken care of at another time. Avoiding the occurrence of the interruption is achievable by changing at least one professional practice (technical and non-technical skills) or changing an organisational axis and it is likely that this can be accommodated by the environment (institutional context, care provided, patient population).	Patient care delivery	Request for administrative information from suppliers (e.g. ambulance drivers); request for information on current orders for health care equipment; request for, or communication of information about a patient or patient care planning that can probably be done at another time (e.g. a patient’s upcoming appointment).
		Patient care planning	Request for information about patient admissions to the department that is likely be available at another time; a student’s lack of knowledge about a protocol to administer medication.
		Care coordination	Interruptions during handovers due to a call from a patient; interruptions by patients during a handover between medical/nursing teams in the corridor; during a handover, a request for further information, or for further details that should probably be known (e.g. homonyms).
		Unit resource coordination	Requests for information about the presence of professionals from other teams in the department, which could probably be obtained at another time; communication of information about patient admissions to the department that is likely be available at another time; search for missing information following the delivery of a drug.
Probably unavoidable	The request or the task that causes the interruption is related to patient safety and probably cannot occur at any other time. Although the interruption can be avoided by changing at least one professional practice (technical and non-technical skills) or by changing an organisational axis, it is unlikely that this can be accommodated by the environment (institutional context, care provided, patient population).	Patient care delivery	Communication or request for information about a patient or internal planning that probably cannot be shared or obtained at another time; request from a patient or their entourage.
		Patient care planning	Passing on information to the family that probably cannot be shared at any other time; communication or request for information about a patient or internal planning that probably cannot be shared or obtained at another time (e.g. no response to a phone call; patient discharge; temporary absence of professionals in the department).
		Care coordination	Interruption by a patient during a handover; communication of external information about a patient’s medical results that probably cannot be shared at any other time; passing on information to the family that probably cannot be shared at any other time; communication or request for information about a patient or internal planning that probably cannot be shared or obtained at another time (e.g. information about a planned admission);
		Unit resource coordination	Request for help with a task that is the team’s responsibility and probably cannot be done at another time (e.g. patient transfer); communication of information to the family that probably cannot be shared at any other time (e.g. documents); communication or request for information concerning a patient or internal planning that probably cannot be shared or obtained at another time.
Unavoidable	The request or the task that causes the interruption is necessary for patient safety and cannot occur at any other time. The interruption cannot be avoided by changing at least one professional practice (technical and non-technical skills) or by changing an organisational axis.	Patient care delivery	Request from a patient or their entourage; emergency management (e.g. pain); management of an unforeseen situation.
		Patient care planning	Communication or request for new or unexpected information about a patient’s health status or their management, which is essential to planned patient care; transmission of medical documents concerning the patient whose care is being planned.
		Care coordination	Call from a patient during a handover; communication of new or unexpected information about a patient’s health status or their management that cannot be shared at any other time;
		Unit resource coordination	Call from a patient; request for help from a second person during a task that is the team’s responsibility, and which cannot be carried out at any other time; Request from a patient or their entourage.

Three other levels of avoidability (‘probably avoidable’, ‘probably unavoidable’, ‘unavoidable’) were derived from this first level (Table 1).

In this study, the term ‘unnecessary’ was used to encompass both ‘avoidable’ and ‘probably avoidable’ interruptions. In this case, it was reasonable to expect the team to be able to devise actions that would improve practices and planning.

Finally, ‘relevant’ interactions between professionals were defined as interactions linking managerial and care delivery functions. They corresponded to interactions involving tasks related to: either (i) unit resource coordination, care coordination and patient care planning or (ii) patient care delivery. Fig 1 illustrates these relevant interactions. Please see Supplementary Material 1 for examples of tasks that illustrate these functions.

Figure 1.

Schematic diagram of relevant interactions between managerial and delivery functions.

managerialcare deliveryotherrelevant interactions (managerial versus delivery)

Observation of interruptions

Observations were conducted from May to September 2019, over the course of a working day for each team. The six professional categories making up a team were observed (administrative, logistical and technical, medical, medicotechnical and psychosocial, paramedical, and management) [20]. A professional was followed for ∼30 min by a pair of observers trained in the methodology. Each of these observers recorded the characteristics of the interruption by hand in a notebook (the Team’IT tool) [20]. The methodology for observing interruptions has been published elsewhere [20]. Supplementary Material 2 outlines the main points (provisional schedule of tasks to be observed, other considerations).

Data collection and processing

Data collected using the Team’IT tool captured what happened during the observation and the characteristics of the professionals observed. They also provided information on the following characteristics of the interruption: (i) the method and duration of the request, (ii) the location of interrupted and interrupting persons, (iii) the reaction of the interrupted person, (iv) the characteristics of the interrupting person, and (v) the classification of interrupted and interrupting tasks according to their function. The items recorded by the Team’IT tool and response modalities are given in Supplementary Material 3.

Data were transcribed the day after observations were carried out and recorded in a dedicated database (Excel®, 2007). The pair of observers compared their notes, reached a consensus if necessary. Data were recorded in the database in order to obtain, among other things, the degree of avoidability of the interruption, and relevant interactions between functions.

The quality of the data was reviewed by examining 460 randomly selected interruptions (20 interruptions/team). The aim was to verify the categorisation, specifically: (i) interrupted and interrupting tasks according to the functions, (ii) how the request was made, and (iii) interrupted and interrupting persons.

Statistical analyses

The number of observed interruptions was calculated for each team, and for all 23 teams. The number of interruptions was calculated as a function of whether they were necessary or not, and/or interactions between managerial and care delivery functions (per team). The frequency of interruptions was expressed as the number of interruptions per professional observed per hour.

Descriptive statistics for each variable were calculated as: (i) position and dispersion indicators (continuous variables) and (ii) totals and corresponding percentages (categorical variables). The profile of observed interruptions was refined by taking into account how necessary they were. Cross-comparisons were run to highlight the evolution of unnecessary interruptions during the working day, for different professional categories, and for the location of the request.

All statistical analyses were performed using R software (version 4.0.0).

Results

Observation characteristics

Table 2a presents variables and descriptive statistics for observations. A total of 286 (potentially) interrupted professionals were observed. Per team, this number was 12.4 (SD = 2.4). The number of occupational categories observed was 3.9 (SD = 0.8). The average number of observed tasks, per team, resulting from the planned schedule was 11.9 (SD = 1.7), and the total number was 274. The duration of observations was 8.1 h (SD = 1.3) per team and 40.9 min (SD = 10.8) per professional.

Table 2.

Variables and descriptive statistics of observations (2A) and of interruptions (2B).

	A
Variables describing observations		n (%) or mean (SD)
Number of professionals observed	Per team	12.4 (SD = 2.4)
Number of occupational categories observed	Per team	3.9 (SD = 0.8)
Number of interrupting categories observed	Per team	6.6 (SD = 1.1)
Number of observed tasks included in the observers’ schedules	Per team	11.9 (SD = 1.7)
	Overall	274 (100)
Classification of observed tasks (included in the observers’ schedules)1 according to function	Unit resource coordination	29 (10.6)
	Care coordination	51 (18.6)
	Patient care planning	49 (17.9)
	Patient care delivery	145 (52.9)
Duration of observations	Per team	486.6 min (SD = 80.4)
	Per professional observed	40.9 min (SD = 10.8)
	B
Variables describing the characteristics of interruptions		n (%) or mean (SD)
Number of interruptions observed	Overall per team	1929 84.3 (SD = 29.9)
Number of unnecessary interruptions observed	Per team	56.1 (SD = 22.5)
Number of interruptions observed linking managerial/delivery functions	Per team	26.1 (SD = 11.5)
Number of unnecessary interruptions linking managerial/delivery functions	Per team	16.5 (SD = 7.7)
Number of interruptions per hour per professional	Per team	10.5 (SD = 3.2)
Number of unnecessary interruptions per hour per professional	Per team	6.9 (SD = 2.5)
Request mode	In person	1132 (58.7%)
	Phone	215 (11.1%)
	Failure/lack of equipment or information	112 (5.8%)
	Medical alarm	71 (3.7%)
	Self-interruption	399 (20.7%)
Duration of the request	≤30 s	1405 (72.5%)
	[30 s; 5 min]	504 (26.0%)
	[5 min; 10 min]	26 (1.3%)
	[10 min; 30 min]	2 (0.1%)
	[30 min; 45 min]	1 (<0.1%)
Reaction of the interruptee	Ignored the question/refused to answer/answered later	92 (4.8%)
	Answered while continuing the task	335 (17.4%)
	Paid attention to the answer	951 (49.3%)
	Stopped	149 (7.7%)
	Don’t know	8 (0.5%)
	Not applicable	394 (20.4%)
Location of the interruptee	Nurses’ station	936 (48.5%)
	Patient’s room	187 (9.7%)
	Corridor	457 (23.7%)
	Other	349 (18.1%)
Characterisation of the interrupter	Medical/paramedical	1295 (67.1%)
	Administrative/logistical and technical/medicotechnical and psychosocial	225 (11.7%)
	Management	125 (6.5%)
	Supplier	40 (2.1%)
	Patient and other	190 (9.8%)
	Not applicable	54 (2.8%)
Avoidability of the interruption	Avoidable	1107 (57.4%)
	Probably avoidable	184 (9.5%)
	Probably unavoidable	241 (12.5%)
	Unavoidable	397 (20.6%)
Classification of the task performed according to the function during an interruption	Unit resource coordination	140 (7.3%)
	Care coordination	688 (35.7%)
	Patient care planning	303 (15.7%)
	Patient care delivery	798 (41.4%)
Classification of the task requested according to the function during an interruption	Unit resource coordination	242 (12.5%)
	Care coordination	300 (15.6%)
	Patient care planning	514 (26.6%)
	Patient care delivery	385 (20.0%)
	Coordination of institutional resources2	28 (1.5%)
	In relation to the institution’s support processes2	13 (0.07%)
	In relation to patient services2	5 (0.03%)
	In relation to the patient’s social life2	77 (4.0%)
	Outside the scope of the professional’s activity	298 (15.4%)
	Don’t know	67 (3.5%)

¹The contact point for the observation teams provided observers with a task schedule (see Supplementary Material 2).

²Tasks carried out by the interrupting professional were categorised into four additional levels, external to the team: coordination of institutional resources; in relation to the institution’s support processes; in relation to patient services; and in relation to the patient’s social life [19, 20].

Interruption characteristics

The data quality review highlighted an interruption categorisation error rate of <5%.

Table 2b presents variables and descriptive statistics for interruptions. A total of 1929 interruptions were observed. Per team, this number was 84.3 (SD = 29.9); the total number of unnecessary interruptions was 56.1 (SD = 22.5); the total number of interruptions involving managerial versus delivery functions was 26.1 (SD = 11.5).

Per team, the average frequency of interruptions was 10.5 (SD = 3.2). This fell to 6.9 (SD = 2.5) when the unnecessary criterion was taken into account. More than half of interruptions were due to a face-to-face request (58.7%). The second-most-common mode was the observed professional interrupting him or herself (20.7%). More than two-thirds of requests lasted ≤30 s (72.5%). Professionals engaged in the response to 49.3% of requests. A total of 6.6 (SD = 1.1) interrupting categories (professionals, and the patient or his or her entourage) were recorded. Medical and paramedical professionals made up the majority of interrupters (67.1%). The patient, or his or her entourage was the origin of 9.8% of requests.

Interruptions were qualified as unavoidable in 20.6% of cases. Statistics for the other categories are: avoidable (57.4%), probably avoidable (9.5%), and probably unavoidable (12.5%). Examples of interruptions by their degree of avoidability are presented in Table 1.

Fig 2a shows how the frequency of interruptions evolved over the working day. It shows the percentage of unnecessary interruptions involving interactions between managerial and care delivery functions. The frequency of interruptions ranged from 4.1 to 16.7, corresponding to the times between 07:00 to 08:00 and 12:00 to 13:00, respectively. The latter period was also the time when the percentage of unnecessary interruptions between managerial and delivery functions was highest (38%).

Figure 2.

Distribution of interruptions during the working day (a), by location observed (b), and by category of interrupter (c).

Fig 2b illustrates the overall frequency of interruptions depending on the location and shows the percentage of unnecessary interruptions involving managerial and care delivery functions. The nurses’ station was the location with the highest observed frequency of interruptions (8.4). (In France, the nurses’ station is a dedicated room that can only be accessed by clinical staff, and is not open to patients.) The percentage of unnecessary interruptions involving managerial and care delivery functions that took place in the nurses’ station was 16%, while 27% took place in the corridor.

Fig 2c presents the frequency of interruptions by category of interrupter. It shows the percentage of unnecessary interruptions involving managerial and care delivery functions. The frequency of interruptions for the different categories was as follows: (i) (para)medical (8.8), (ii) administrative/medicotechnical/logistical (1.8), and (iii) suppliers (0.4). The percentage of unnecessary interruptions involving interactions between managerial and care delivery functions was similar for the first two categories (22%), and 54% for the latter. The patient and his or her entourage interrupted the professional on average 1.8 times per hour.

Discussion

Statement of principal findings

This French study is part of the research stream that seeks to ensure care safety by studying the interruptions that affect professionals during their working day. It adopts the definition of an interruption currently used in France [18], and characterises interruptions by considering their typology, duration, the interrupter, the reaction of the professional interrupted, and the frequency of interruptions. The study found that the average number of interruptions per hour and per professional (for all categories) was 10.5, and this figure is higher than that usually quoted by French health authorities (6.7) [18]. This frequency was reported by Biron et al. in a review of other countries (2622 interruptions; 402.5 h of observations of nurses during drug-related tasks) [24]. In our study, the majority of interruptions were face-to-face (58.7%) and to a lesser extent by telephone (11.1%). The percentage of interruptions attributable to the patient (9.8%) was smaller than the number attributable to healthcare professionals. Concerning the reaction of the interruptee, 49.3% stopped doing what they were doing to pay attention to the request. These findings complement those obtained in another French study, conducted in 2015 [18].

Our study makes the following contributions: (i) it considers interactions between functions that could increase safety of care risk; (ii) it defines, for the first time, an ‘avoidable’ interruption; and (iii) it reports, for the first time, the percentage of unnecessary interruptions. The percentage of interactions between managerial and care delivery functions (31.3%), highlighted here, is similar to that reported in McCurdie et al. (33.3% in an audited intensive care unit) [19]. Although 20.6% of interruptions are unavoidable, 66.4% are unnecessary. The percentage of unnecessary interruptions involving interactions between managerial and care delivery functions is concentrated in the morning. More specifically, the percentage is two times higher when the interrupter is a professional in the (para)medical or administrative/medicotechnical/logistical categories compared to the patient. Finally, the corridor is the place where this type of interruption is most likely to happen (twice as frequently as in the nurses’ station). These results show where teams have room for improvement, in reducing the number of interruptions affecting the safety of care.

Strengths and limitations

The present study is the first of this magnitude conducted in the inpatient hospital sector. One limitation concerns the data transcription. It became apparent that the variable ‘nature of the requested task’ would be the most complex to categorise (results are not given here), and that this would have a direct impact on outcome variables concerning interactions between managerial and delivery functions. However, the examples used to illustrate each function and avoidability criterion are carefully selected and should help future observers to grasp the definition of avoidability. Finally, the research targeted interruptions that occurred during the day. Their characteristics cannot be directly transposed to interruptions occurring at night.

Interpretation within the context of the wider literature

A comparison of our data with those of other large-scale studies remains a challenge [7], due to differences in the investigated sectors, methodology, and initial objectives [2, 13, 25]. Nevertheless, our study makes a novel contribution to the literature—although it should be noted that demonstrating the link between an interruption and the occurrence of an adverse event associated with care remains difficult [25–27]. Our observers assessed interactions between work functions, and investigated the avoidability of interruptions with reference to the need, at a given moment, for the interruption, in the interests of patient safety.

It should be noted that the interruptions that we describe as ‘avoidable’ should be contextualized with respect to the work environment (taking into account the operational system and the need for efficiency) as this environment influences whether an individual will interrupt [27]. Our observations were a first step in helping teams to think about how to manage interruptions. Subsequently, participants were invited to analyse their practices in order to contextualize the observed interruptions. The final objective was for them to be able to identify potential improvements to their own operational system, and associated safety barriers. Among the various solutions, some sought to improve communication and cooperation between professionals by deploying secure communication tools. The aim of the latter was to provide professionals with ritualized, effective spaces, and shared working methods. As the project was carried out in the context of a COVID crisis, it was not possible to assess the impact of these solutions on the characteristics of interruptions. Furthermore, as teams initially limited their solutions to ways to prevent the occurrence of interruptions, further efforts need to be made to not only block and recover from interruptions, but also mitigate their effects (results not shown).

The literature highlights that interruptions can contribute to errors by increasing the probability that certain tasks will be forgotten or delayed [26, 28]. However, interruptions can be beneficial [29]. In this context, there is no standard definition of the relevance of interruptions. One way forward could be to assess the latter with reference to when the benefit of the interruption outweighs the consequences of the risk incurred. An aspect that merits further attention is the organizational or clinical value of the information transmitted by the person interrupting or being interrupted [28]. It would even also possible, for example, to define value from the point of view of the interrupting patient as well as the professional, in cases where the patient experience is the fundamental criterion for the definition of quality [29].

Implications for policy, practice, and research

Our results can be capitalised on in several ways. Applied to care teams, this includes the following actions: (allowing themselves) to postpone an avoidable interruption; thinking about how to protect tasks that take place between 12:00 and 13:00; improve the ergonomics of the nurses’ station; give special attention to tasks that take place in the patient’s room; and making medical and nursing professionals more aware of the importance of interruptions. Too many misconceptions remain, notably those concerning the patient’s entourage, suppliers, or the telephone, all of which are said to be the source of numerous interruptions.

Conclusions

This study highlighted the characteristics of interruptions of professionals working in the inpatient hospital sector. An average of 10.5 interruptions were recorded per hour per professional, 57.4% of which were avoidable. The period between 12:00 and 13:00 emerged as the riskiest in terms of safety of care. Care teams should focus on: the making medical and nursing professionals more aware of the issue of interruptions, and each team could decide how to best-mange interruptions, in the context of their working environment.

Supplementary data

Supplementary data is available at International Journal For Quality In Health Care online

Funding

This work was supported by the Direction Général de l’Offre des Soins (DGOS) [PREPS 18-0047/, 201].

Data availability statement

The authors agree to share the data. Please contact the corresponding author should you require any of the results from this study.

Author contributions

Delphine Teigné and Lucie Cazet (data collection, interpreted the data and drafting, revision, and design); Lucie Cazet (data collection, interpreted the data and drafting, and revision); Guillaume Mabileau (statistical analysis, interpreted the data and drafting, revision, and design); Noémie Terrien (interpreted the data and drafting, revision, and design); and Jean-Claude Maupetit, Gabriel Birgand, and Leila Moret (revision and design).

All authors have read and approved the manuscript.

Ethics and other permissions

The IMPACTT non-interventional research protocol was authorised by the Health Research Information Processing Advisory Committee and received approval from an ethics committee (the Gneds) on 17 January 2019. Information about the project was posted on the premises of the observed teams. Participation (interruptees and interrupters) was conditional on oral consent being given. In accordance with articles L1121-1 and R1121-2 of the French Public Health Code, Institutional Review Board approval was not necessary.

Consent for publication

Not applicable.

Trial registration

NCT03786874 (26 December 2018).