Family-centered rounds is an approach to patient care that includes patients and their family members as part of the daily discussion of medical decision-making, and often also includes practitioners of supporting disciplines.1,2 This type of rounding has become increasingly popular over the past decade, as it has been shown to increase family satisfaction, improve team communication, streamline management, and reduce medical errors.3,4 Although bedside rounding lets the attending physician explain management decisions to the family, it also informs the multidisciplinary team in a collaborative manner.
The pediatric intensive care unit (PICU) at Randall Children's Hospital, a 24-bed regional referral unit that handles more than 1,000 medical and surgical admissions yearly, has rounded with a large multidisciplinary team for the past 15 years and has received consistent support as a result of family and staff satisfaction. However, we identified several barriers to rounding efficiency during family-centered rounds. One barrier was the steep learning curve associated with the implementation of the electronic medical record system.4 We have since employed scribes, who help physicians by creating electronic charts for each patient, detailing the daily plan before rounding, and editing charts during rounds as treatment plans change. Physician assistants (PAs) were hired in 2014 to assist with admitting and transferring patients, consulting with providers, handling complex order entry, and performing procedures.
We round with a large multidisciplinary team, as integrative teamwork has been shown to relate directly to improved clinical performance.5 However, the size of the patient rooms became a barrier because they cannot easily accommodate a large number of people.4 We instituted hallway rounds (Figure 1) to ensure sufficient space for the entire team and family members. Hallway rounds also let the team overcome the issue of isolation precautions by reducing the number of people entering and exiting the room of an isolated patient.
A third barrier to rounding efficiency was frequent interruptions during rounds. For the purpose of this study, we defined an interruption as any event that caused discontinuity in the physician's patient presentation. A verbal interruption can significantly lengthen the time necessary to discuss a patient's care, extending overall rounding time and decreasing efficiency.1 Interruptions and other distractions also may pose a threat to cognitive decision-making, as a switch between two cognitive tasks increases errors.6 Furthermore, an interruption as short as 10 seconds can cause people to forget exactly what they were doing during their previous task.7
We evaluated the effect of the presence of a PA on reducing interruptions during family-centered rounds in the PICU, and found that the presence of a PA significantly reduced the number of interruptions during rounds. Reducing interruptions decreases the cognitive burden from task switching and may inherently reduce error and improve patient safety. We are particularly interested in how PAs can be used in the treatment of critically ill patients in emergency and trauma units that are particularly susceptible to interruptions and multitasking behavior.
We conducted daily family-centered rounds in the Randall Children's Hospital PICU, a 24-bed regional referral unit providing around-the-clock attending physician coverage of complex medical and surgical patients. Family-centered rounds use a multidisciplinary team consisting of a pediatric intensivist, PA, pharmacist, nutritionist, charge nurse, social worker, respiratory therapist, bedside nurse, scribe, and patient's family. Rounding occurs in order of room number. After a system-based review of each patient's case, including addressing clinical bundles and allowing for family questions, decisions about the patient's care are made and necessary orders placed. As we do not have trainees on our unit, the PA helps place orders, which lets the physician address questions or other issues raised by the staff and family members. Once this is complete, the team moves as a whole to the next patient.
Rounding times and interruptions
During family-centered rounds conducted on 2,657 patients between May 2014 and October 2015, scribes recorded the total time spent on each patient, the total rounding time (using a digital time recorder), occurrence of interruptions, and whether a PA and the patient's family were present. The timer was started when the presentation began on the first patient and stopped when the team moved to the next patient. Time spent walking from room to room was included in the time spent on the next patient. The average distance between rooms was 23.5 ft and it took an average of 5.8 seconds (range, 3 to 14 seconds) to cover this distance.
Interruptions that occurred during patient presentations were noted and the associated duration was added to the patient's individual rounding time on a paper data collection sheet designed by the senior author (MTB). The type of interruption (consult requests, pages, emergent procedures, and request for new admissions) also was recorded by the scribe on the same data sheet. These data were transferred to an electronic database by the scribe after the completion of rounds (Table 1). The paper logs were stored in chronologic order.
For weekday rounding encounters, data concerning interruptions were tabulated to examine the proportion of rounding encounters that were interrupted when a PA was present or absent. Initial univariate logistic regression analyses were performed to determine which predictors (presence of a PA, presence of family, intubated or tracheostomy-dependent patients) were significantly associated with the occurrence of interruptions. Multivariable logistic regression was then conducted, including all three influential variables (presence of a PA, presence of family, intubated or tracheostomy-dependent patients). All statistical analyses were conducted using the R Language for Statistical Computing.8
This study was granted institutional review board exemption because it did not directly affect human subjects and all information was deidentified at the time of collection.
When a PA was not present during rounding, 178 of 1,137 (15.7%) encounters were interrupted. However, when a PA was present, 163 of 1,520 (10.7%) encounters were interrupted. This represents a 31.8% lower rate of interruptions when a PA was present. In a univariate logistic regression model that treated all rounding encounters from each physician as being clustered, the presence of a PA during rounds was significantly associated (P < .001) with a 35.3% lower likelihood of an interruption occurring. In other univariate models, neither the presence of family members during rounding (P = .49) nor an intubated state (P = .22) was significantly associated with the likelihood that a rounding encounter would be interrupted. In a multivariable logistic regression model that treated all rounding encounters from each physician as being clustered and that controlled for all three influential variables (presence of a PA, presence of family, intubated or tracheostomy-dependent patients), once again only the presence of a PA during rounds was significantly associated (P < .001) with a 35.4% lower likelihood of an interruption occurring. See Table 2 for a tabulation of coefficients and P values from the univariate and multivariable logistic regression models constructed.
The duration of the interruption was not captured. For uninterrupted encounters the median length of rounding duration was 6:38 minutes (first quartile, 4:20 minutes; third quartile, 9:46 minutes), whereas for interrupted encounters the median length of rounding duration was 9:50 (6:58 to 13:54) minutes. We can therefore deduce that, on average, an interruption added about 3:12 minutes to the duration of the encounter (48% longer). Similarly, when we examined the rounding encounter duration for encounters that did and did not have a PA present, we found that having a PA present on average shortened the rounding encounter duration by about 44 seconds (about 10% shorter: with PA duration = 6:42 [4:17 to 9:50] minutes; without PA duration = 7:26 [4:54 to 11:12] minutes).
We found that the interruption types fell into five different groupings. The most common were general discussions and updates, including off-topic conversations that delayed a patient encounter. The second most common type included electronic interruptions, such as phone calls and pages. Mechanical failure (such as computer malfunction) and waiting for staff were almost equal in frequency and rather uncommon. The least common interruption type was emergent procedures such as CPR, intubation, or emergent admission. From these findings, we conclude that a PA would be best used in addressing the most common interruptions, such as phone calls, pages, and receiving updates.
Using a detailed timing method during family-centered rounds, we found that the presence of a PA decreased interruptions by 31.8% using tabulated data and by 35.4% via multivariable logistic regression analysis (P < .001). By covering the ICU during the 2-hour window of multidisciplinary rounds, the PA let the attending have uninterrupted interactions with families and other providers. Outside of rounds, PAs perform arterial line insertions, lumbar punctures, intubations, and manage ventilators. Given that we do not have fellows or residents, PAs serve a vital role in complex patient care.
Interruptions during high-risk activities have been shown to increase error rates.9,10 A study conducted in an Australian hospital found that providers were interrupted as frequently as every 2 minutes, with each interruption resulting in a 12.7% increased risk of clinical error.10 Interruptions not only affect cognitive abilities in the medical field but also in other high-risk areas, including aviation. The sterile cockpit, a type of no-interruption zone for pilots, was instituted by the Federal Aviation Administration in 1981. This act ensures that pilots are not interrupted by nonessential tasks during flight operations below 10,000 ft. Distractions during takeoff and landing accounted for 72% of 76 reported airline incidents (an occurrence, other than an accident, that affects or could affect the safety of operations).9 Reducing cognitive load by limiting interruptions reduces the risk of error, and inherently increases patient safety.
Multiple interruptions, which can lead to switches in cognitive tasks, may increase error rates and thus pose a threat to patient safety. Interruptions not only affect the person responding to the interruption (typically the physician) but also the multidisciplinary team as a whole given that the physician typically leads the discussion. The effect of interruptions during a high-risk medical activity, such as prescribing and administering medications, has spurred the creation of no-interruption zones.9 Interruptions while administering medications caused a 13% increase in clinical errors among nurses; hospitals using no-interruption zones observed a 20% reduction in medication administration errors.9,11 The creation of the no-interruption zone may encourage mindfulness among the staff, reducing error rates and promoting patient safety. Staff mindfulness is a growing topic of discussion in the effort to reduce interruptions in the healthcare setting.10
Families who have participated in family-centered rounds have reported anecdotally that they are there to be informed, to participate in their child's care, and to fulfill their parental role. They considered frequent distractions and interruptions as a lack of common courtesy. While the attending is answering phone calls and others are speaking among themselves, family members find it difficult to know upon whom to focus.12
Some of the most common interruptions are electronic, and include phone calls or pages (Table 3). We encourage consulting teams to speak with the PA when a physician is rounding on a patient. In this way, we hope to reduce the number of nonurgent, face-to-face interruptions.
Furthermore, we are creating a no-interruption time during family-centered rounds, which to our knowledge has not yet been reported. Additionally, we encourage our multidisciplinary providers to write nonurgent questions down to ask at more appropriate times than during the patient presentation. Thus, we aim to create mindfulness among our staff in order to foster a focused and structured environment that helps to decrease unnecessary interruptions.10 By creating a methodology in which interruptions during family-centered rounds are minimized, we are attempting to decrease the physician's cognitive burden of switching between competing tasks. Ultimately, we believe that the employment of a PA to manage predictable interruptions and distractions may safeguard patients.
In the past decade, the PA scope of practice has been expanded to include the care of hospitalized patients in many areas, such as neonatal, critical care, emergency services, and trauma care.13-25 Emergency and trauma care may be more susceptible to interruptions and subsequent errors due to the number of providers and consultants, availability and limitations of resident staffing, patient instability, and rapid decision-making requirements.26 In our PICU, PAs play a vital role in complex pediatric patient care and have become essential members of our medical team. Whether their incorporation in medical teams is associated with efficiency and improved patient outcomes is an area that requires further research.
This study had several limitations. PAs were present during family-centered rounds only on weekdays. The changes that occur over the weekend (staffing limitations, lack of routine surgical admissions) may affect interruption rate, and may or may not be affected by a PA's presence. Furthermore, we can extrapolate from our data that there will be a certain frequency of interruptions due to events that a PA cannot manage independently. For example, new admissions must be accepted by the physician. The attending physician must perform emergent procedures that are outside the PA's scope of practice within that unit. Also, some interruptions are unavoidable, such as computer malfunctions; however, by employing a method to manage interruptions, such as using a PA to respond to the most common distractions, workflow and team communication can improve. Another possible limitation is human error while recording rounding times. A scribe may have started the timer late, or made an error while recording the result. This limitation is unpredictable, and is unlikely to have biased the outcome in one direction or another.
Finally, as our PICU's multidisciplinary team does not include resident or fellow trainees, these results may or may not apply to programs with trainees, who would fulfill some of the same duties as the PAs in our practice.
Although family-centered rounds increase family, patient, and staff satisfaction, rounding efficiency remains a continual challenge in the PICU. We have demonstrated that rounding efficiency can be significantly improved by using a PA to respond to interruptions. Further research is needed to determine whether patient outcomes are improved by more efficient and streamlined rounding, due to the decreased need for the physician to switch between cognitive tasks. By creating a no-interruption time during family-centered rounds, we expect clinical error rates to decrease. We also anticipate increased family and staff satisfaction with this focused and more organized approach to multidisciplinary rounding.
1. Bhansali P, Birch S, Campbell JK, et al. A time-motion study of inpatient rounds using a family-centered rounds model. Hosp Pediatr
2. Sisterhen LL, Blaszak RT, Woods MB, Smith CE. Defining family-centered rounds. Teach Learn Med
3. Mittal V. Family-centered rounds: a decade of growth. Hosp Pediatr
4. Gonzalo JD, Heist BS, Duffy BL, et al. Identifying and overcoming the barriers to bedside rounds: a multicenter qualitative study. Acad Med
5. Manser T. Teamwork and patient safety in dynamic domains of healthcare: a review of the literature. Acta Anaesthesiol Scand
6. Rogers R, Monsell S. Costs of a predictable switch between simple cognitive tasks. J Exp Psychol Gen
7. Beyea SC. Distractions, interruptions, and patient safety. AORN J
8. R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. www.R-project.org
. Accessed March 1, 2018.
9. Anthony K, Wiencek C, Bauer C, et al. No interruptions please: impact of a no interruption zone on medication safety in intensive care units. Crit Care Nurse
10. Beyea S. Interruptions and distractions in health care: improved safety with mindfulness. https://psnet.ahrq.gov/perspectives/perspective/152
. Accessed March 1, 2018.
11. Magrabi F, Li SY, Dunn AG, Coeira E. Challenges in measuring the impact of interruption on patient safety and workflow outcomes. Methods Inf Med
12. Levin AB, Fisher KR, Cato KD, et al. An evaluation of family-centered rounds in the PICU: room for improvement suggested by families and providers. Pediatr Crit Care Med
13. Oswanski MF, Sharma OP, Raj SS. Comparative review of use of physician assistants
in a level I trauma center. Am Surg
14. Silver HK, McAtee P. Speaking out. Should nurses substitute for house staff. Am J Nurs
15. Committee on Hospital Care. Policy statement. Role of physician assistant in the care of hospitalized children (RE9864). Am Acad Pediatr
16. Carzoli RP, Martinez-Cruz M, Cuevas LL, et al. Comparison of neonatal nurse practitioners, physician assistants
, and residents in the neonatal intensive care unit. Arch Pediatr Adolesc Med
17. Dubaybo BA, Samson MK, Carlson RW. The role of physician-assistants in critical care units. Chest
18. Goldfrank L, Corso T, Squillacote D. The emergency services physician assistant: results of two years' experience. Ann Emerg Med
19. Jones PE, Cawley JF. Physician assistants
and health system reform. Clinical capabilities, practice activities, and potential roles. JAMA
20. Riportella-Muller R, Libby D, Kindig D. The substitution of physician assistants
and nurse practitioners for physician residents in teaching hospitals. Health Aff (Millwood)
21. Schulman M, Lucchese KR, Sullivan AC. Transition from housestaff to nonphysicians as neonatal intensive care providers: cost, impact on revenue, and quality of care. Am J Perinatol
22. Sox HC Jr, Ginsburg JA, Scott HD. Physician assistants
and nurse practitioners. Ann Intern Med
23. Sturmann KM, Ehrenberg K, Salzberg MR. Physician assistants
in emergency medicine. Ann Emerg Med
24. American College of Emergency Physicians. Guidelines on the role of physician assistants
in emergency departments. Ann Emerg Med
25. Miller W, Riehl E, Napier M, et al. Use of physician assistants
as surgery/trauma house staff at an American college of surgeons-verified Level II trauma center. J Trauma
26. Sen S, Kranzler HR, Didwania AK, et al. Effects of the 2011 duty hour reforms on interns and their patients: a prospective longitudinal cohort study. JAMA Intern Med