Published online Sep 9, 2026. doi: 10.5492/wjccm.122427
Revised: May 26, 2026
Accepted: June 5, 2026
Published online: September 9, 2026
Processing time: 131 Days and 14.8 Hours
Early mobility in the pediatric intensive care unit (PICU) is safe and associated with improved cognitive and functional outcomes while reducing complications and hospital length of stay. Protocolized early mobility programs guide activity based on physiologic criteria and increase mobilization rates. However, critically ill children remain under-mobilized. Nurses, key agents of mobility imple
To evaluate discrepancies between nurse-reported mobility, electronic health record documentation, and early mobility protocol expectations in the PICU.
We conducted an observational study evaluating nursing-led mobility in 101 patients admitted ≥ 3 days to a large, academic PICU. Key variables included physiologic mobility level (1 = most restrictive to 3 = most liberal), the highest level of mobility (HLM) achieved, and the number of mobilizations. Data were obtained from the electronic health record, end-of-shift nurse interview [registered nurse report (RN-report)], and compared with protocol-expected mobility using Cohen’s Kappa. Firth penalized logistic regression assessed age, pediatric risk of mortality score, and pediatric cerebral performance category as predictors of discordance.
Agreement between protocol-expected and RN-reported mobility level was moderate (53.5%; κ = 0.32) and agreement between protocol-expected and RN-reported HLM was poor (40.6%; κ = 0.20). In univariate analysis, mechanical ventilation [odds ratio (OR) = 0.34, 95% confidence interval (CI): 0.19-0.97, P = 0.016], vascular access lines (OR = 0.43, 95%CI: 0.13-0.69, P = 0.043), and sedation > 30 minutes (OR = 0.30, 95%CI: 0.13-0.69, P = 0.005) were associated with reduced odds of HLM discordance. Conversely, severe disability at baseline was associated with higher odds of having a discordant mobility level (OR = 8.33, 95%CI: 2.50-27.76, P = 0.001), and a discordant HLM (OR = 3.71, 95%CI: 1.15-12.01, P = 0.029), even after adjusting for age and illness severity.
In a longstanding PICU mobility program, discordance exists between protocol-expected and RN-reported mobility. Severe baseline disability increases this discordance, highlighting the need to evaluate the fidelity of early mobility programs.
Core Tip: This initiative found substantial discordance between protocol-expected, health record documented, and nurse-reported mobilizations in a pediatric intensive care unit with a 12-year early mobility program. Nursing uncertainty re
- Citation: Manikandan D, Lenker H, Mennie C, Morgenstern S, Furniturewala S, Hwang L, Hajnik K, Brown KM, Shilkofski N, Kudchadkar SR, LaRosa JM. Discordance between early mobility protocol expectations and nurse-led mobilizations of critically ill children: A quality improvement initiative. World J Crit Care Med 2026; 15(3): 122427
- URL: https://www.wjgnet.com/2220-3141/full/v15/i3/122427.htm
- DOI: https://dx.doi.org/10.5492/wjccm.122427
In the last two decades, mortality rates in pediatric intensive care units (PICUs) have reached an all-time low of nearly 2%. Despite this achievement, nearly one-third of patients who survive the PICU are discharged with a poorer functional status compared to the time of admission[1-3]. Acquired morbidities, including supplemental nutrition needs, motor deficits, and chronic respiratory support, may persist for years after a PICU stay and are common among patients with neurological conditions and those requiring mechanical ventilation[4,5]. Early mobilization of patients in the ICU has been demonstrated to be safe and improve cognitive function and the ability to perform activities of daily living after discharge, while reducing the risk of respiratory complications and length of the hospital stay[6,7].
Early mobility protocols provide age and developmentally appropriate activity guidelines for critically ill children. These protocols assign mobility levels based on physiologic criteria (e.g., intubation status, vasoactive infusions) and recommend activities appropriate for each level[8-11]. Quality improvement studies have demonstrated that the initiation of these protocols results in an absolute increase in the number of mobilizations and a decrease in perceived barriers to mobility by PICU staff[11-13]. However, despite the evidence, rehabilitation consultations and mobility sessions remain underutilized in this population[14].
Many studies have examined the barriers to early mobility in both pediatric and adult populations; however, only one adult study and no pediatric studies have evaluated actual deviations from an early mobility protocol[15,16]. The adult study showed that patients who were clinically appropriate for mobility were not mobilized on 22% of study days[17]. Such deviations may reflect misclassification of mobility levels or under-mobilization of otherwise eligible patients.
To investigate this phenomenon, we conducted a quality improvement project to increase nurse-led mobility among patients admitted to the PICU for ≥ 3 days. Our primary aim was to evaluate discrepancies between achieved and pro
We conducted a quality improvement initiative aimed at increasing nurse-led mobilizations in critically ill children using in situ simulation[18]. This project was designated as a quality improvement initiative by our institutional review board and was reviewed and acknowledged as exempt, No. IRB00289007.
PICU Up! is a structured, interdisciplinary early rehabilitation and mobility program designed to improve outcomes for critically ill children[11]. It involves training PICU staff and implementation of a tiered age-appropriate activity plan based on strict clinical parameters (Figure 1). Patients are assigned a PICU Up! mobility level from 1 to 3 based on phy
The cohort included 101 patients aged 1 day to 17 years who had been admitted to the Johns Hopkins Hospital PICU for ≥ 72 hours and ≤ 7 days as of 7:00 AM on the day of data collection. Data were collected from February 2024 to October 2024. Patients with extracorporeal membrane oxygenation, open chest or abdomen, unstable fractures, or medical orders specifying a contraindication to mobility were excluded.
The primary outcomes were the discrepancy between the electronic health record (EHR) documented, RN-reported , and expected PICU Up! levels, number of mobilization events per patient per shift (7a-7p of the study day), the highest level of mobility achieved (HLM), and the occurrence of adverse events. HLM was broadly categorized based on the maximal activity that the patient achieved per shift (Figure 1). Passive in-bed activities (e.g., bedrest, range of motion, or passive movements) were reserved for patients assigned to PICU Up! level 1 - active in-bed activities (e.g., exercise in bed or sitting at the edge of the bed) were expected of patients identified as level 2. Out-of-bed activities (e.g., being held by caregivers or sitting in a chair) were expected for level 3 patients. The final category was standing or ambulating, and this was expected of all level 3 patients who had the ability to perform these tasks at baseline.
Observed outcomes were captured using two methods: (1) A 3-minute to 5-minute in-person interview with the patient’s nurse conducted at 18:00 (1 hour before the end of the shift) on the study day (RN-report); and (2) Data extracted from the EHR. In the primary quality improvement initiative, RN-report was used as the primary outcome metric because, at the time, documentation of mobility in the EHR was not standardized in the unit[18]. Specifically, nurses were asked to report the patient’s PICU Up! level, HLM, total number of mobilizations, and any adverse safety events that occurred during the shift. An unknown level was marked if, during the post-shift interview, the nurse was unable to map the patient’s clinical characteristics to the protocol-corresponding PICU Up! level and the highest level of mobility. In these cases, the interviewer presented the nurses with a visual aide of the protocol, and an “unknown level” indicated that nurse uncertainty persisted despite use of this aide. For regression modeling, we used RN-report as a proxy for bedside care delivery. This metric was selected based on feasibility, as continuous direct observations of mobility were not feasible; it represents a surrogate measure of mobility practices and is subject to reporting biases.
The expected outcomes were derived retrospectively from the EHR using the PICU Up! protocol criteria (Figure 1) by the primary author, Manikandan D. Each physiologic criterion outlined in “step 1: Screening” in Figure 1 was captured on a data collection form. A mobility level was then assigned based on the parameters for inclusion criteria as per the protocol. Uncertain physiologic inclusion criteria were adjudicated by the author, LaRosa JM, which occurred in less than 20% of patients. Adjudication most often occurred due to incomplete or contradictory information in the EHR or the need for clinical expertise above the level of the primary reviewer. The authors were not blinded to the EHR-documented or RN-reported PICU Up! level achieved.
Demographic and clinical information was obtained from the EHR and included age, sex, admission reason, Pediatric Cerebral Performance Category (PCPC) Score, quantity and nature of medical equipment, respiratory support status, continuous sedative use for at least 30 minutes in the last 24 hours, vasoactive drug administration in the last 24 hours, delirium screening status, and Pediatric Risk of Mortality Score III (PRISM)[19,20]. Details of the respiratory support, such as fraction of inspired oxygen (FiO2) and positive end-expiratory pressure, were also collected.
Categorical variables were reported as percentages and counts, and continuous variables were reported as means and standard deviations. Medians and interquartile ranges (IQR) were used if the latter had a non-normal distribution. Missingness across the data was < 5%. Agreement between the PICU Up! level in the EHR, RN-report, and the expected outcome for each patient was compared with absolute agreement (%) and Cohen’s Kappa statistic (κ)[21,22]. Due to our small sample size, a Firth Penalized logistic regression was used to calculate odds ratios (ORs), with 95% confidence interval (CI) to identify factors associated with (1) Discordance between the expected and RN-reported PICU Up! levels; and (2) Discordance between expected and RN-reported HLM. Covariates were defined a priori, consistent with past PICU Up! research and included age category (0-2, 3-6, 7-12, 13-18, or > 18 years of age); PCPC score (good, mild disabi
The dataset included a total of 101 patients. The mean age was 7.10 ± 5.70 years. The sample was predominantly male (58.4%; 59/101) with a median PICU length of stay of 5 days (IQR = 4-6). The most common reason for PICU admission was a medical admission (76.2%; 77/101), followed by surgical admission (13.9%; 14/101), and trauma or other (9.9%; 10/101). Illness severity was measured by the PRISM score, and most patients (88.1%; 89/101) had a low-risk score of 0-10, with the remainder classified as moderate risk or greater (11.9%; 12/101). Baseline functional status was assessed by the PCPC Score, and 49.5% (50/101) of patients had good baseline function, whereas 22.8% (23/101) had severe disability. Most patients (53.5%; 54/101) required no respiratory support or supplemental oxygen through low flow nasal cannula, while the remainder received support through non-invasive positive pressure ventilation and tracheostomy collar (11.9%; 12/101) or invasive mechanical ventilation (34.7%; 35/101). The most common piece of medical equipment was gastric/nasogastric tubes (66.3%; 67/101), and more than half the sample had invasive vascular access such as arterial and central venous lines (53.5%; 54/101). Nearly a third of patients received sedation lasting > 30 minutes in the past 24 hours (34.7%; 35/101), and a minority of the sample (4.2%; 4/101) were documented as experiencing delirium in the past 24 hours (Table 1).
| Characteristic | |
| Age | |
| < 2 years | 27 (30.7) |
| 3-5 years | 12 (13.6) |
| 6-12 years | 30 (34.1) |
| 13-18 years | 19 (21.6) |
| Gender | |
| Male | 59 (58.4) |
| Female | 42 (41.6) |
| PICU day (median, IQR) | 5 (4-6) |
| Reason for PICU admission | |
| Surgical | 14 (13.9) |
| Medical | 77 (76.2) |
| Trauma or other | 10 (9.9) |
| PRISM score | |
| Low risk (0-10) | 89 (88.1) |
| Moderate risk or greater (11-30+) | 12 (11.9) |
| Preadmission PCPC score | |
| Good | 50 (49.5) |
| Mild disability | 15 (14.9) |
| Moderate disability | 13 (12.9) |
| Severe disability/coma or vegetative state | 23 (22.8) |
| Respiratory support category | |
| No support or cannula/mask | 54 (53.5) |
| NIPPV/tracheostomy collar | 12 (11.9) |
| Invasive mechanical ventilation | 35 (34.7) |
| Equipment | |
| Airway equipment (endotracheal tube/tracheostomy) | 36 (35.6) |
| Lines (arterial line or central venous line) | 54 (53.5) |
| Drains and catheters | 28 (27.7) |
| Tubes (gastric/nasogastric tube) | 67 (66.3) |
| Any sedation > 30 minutes | 35 (34.7) |
| Any vasoactive within 24 hours | 10 (9.9) |
Across all three data sources - EHR, nurse end-of-shift interview, and protocol-derived expected values - most patients were classified as PICU Up! level 3 (EHR: 47.5 %; 48/101, RN-report: 50.5%; 51/101, expected: 53.6%; 54/101). By EHR documentation, 12.9% (13/101) of patients were assigned to PICU Up! level 1 and 8.9% (9/101) by RN-report, compared with 23.8% (24/101) expected (Table 2).
| PICU Up! level | EHR1 | RN-report2 | Expected |
| Level 1 | 13 (12.9) | 9 (8.9) | 24 (23.8) |
| Level 2 | 26 (25.7) | 13 (12.9) | 23 (22.8) |
| Level 3 | 48 (47.5) | 51 (50.5) | 54 (53.5) |
| Undocumented/unknown | 14 (13.9) | 28 (27.7) | 0 (0) |
| Highest level of mobility achieved | EHR3 | RN-report4 | |
| Bedrest or passive movement | 54 (53.5) | 34 (33.7) | 28 (27.7) |
| In-bed activities | 8 (7.9) | 14 (13.9) | 19 (18.8) |
| Out-of-bed activities | 14 (13.9) | 31 (30.7) | 24 (23.8) |
| Standing or ambulating | 25 (24.8) | 22 (21.8) | 30 (29.7) |
| Number of nurse-led mobilizations per shift (median, IQR) | 1 (2) | 6 (2) | 6 (0) |
| Potential safety event rate per mobilization | 0.02 (2/101) | 0.02 (12/606) | 0 (0) |
Notably, the PICU Up! level was undocumented in 13.9% (14/101) of EHR records and reported as unknown in 27.7% (28/101) of RN-reports. Of the 28 patients classified with an unknown level by the nurse, 15 were expected to be level 3 per protocol, and yet 16/28 were reported to have an HLM of passive range of motion or in-bed activities, consistent with level 1 or level 2 mobility. Overall, agreements between expected and EHR-documented PICU Up! level (56.4%; κ = 0.34), as well as agreement between expected and RN-reported PICU Up! level were similarly modest (53.5%; κ = 0.32) (Table 2, Supplementary Table 1).
With respect to the HLM achieved, 53.5% (54/101) of patients were documented in the EHR as performing passive in-bed activities, compared to 33.7% (34/101) per RN-report. These numbers are higher than the 27.7% (28/101) of patients who were expected to have passive in-bed activities by protocol. Similarly, active in-bed activities were documented by EHR to have occurred in 7.9% (8/101) of patients and RN-reported in 13.9% (14/101) of patients, although they were expected in 18.8% (19/101) of patients by protocol. In contrast, out-of-bed activities were reported in 30.7% (31/101) of patients by RN-report, but only expected in 23.8% (24/101) and EHR-documented for 13.9% (14/101) of patients.
The median number of mobilizations per shift was 1 (IQR = 2) in the EHR, compared to 6 (IQR = 2) by nursing end-of-shift reports and 6 expected. Overall, these data are supported by a generally poor agreement between the expected and EHR documentation of HLM (39.6%; κ = 0.17) as well as RN-report of HLM (40.6%; κ = 0.20) (Table 2, Supplementary Table 1). Safety events with mobilization were rare (median = 0.02 potential safety events per mobilization) across EHR and RN-report , and there were no equipment dislodgements or cardiac arrests (Table 2)[6,7].
Discordance between RN-reported and protocol-expected PICU Up! level was present in 46.5% of cases (47/101). Discordance in HLM was similarly high at 59.4% of patients (60/101), with a lower level of mobilization most common in patients who were expected to be ambulatory (Supplementary Table 2). Univariate logistic regression identified clinical and demographic factors associated with the discordance between RN-report and expected protocol-assigned values for both PICU Up! mobility level and the highest level of mobilization achieved (Table 3). Severe disability at baseline was associated with higher odds of having a discordant PICU Up! level compared with good baseline function (OR = 8.29, 95%CI: 2.56-26.91, P < 0.001), whereas mild or moderate disability was not. ICU admission reason, respiratory support, equipment, the presence of delirium, and patient demographic factors were not associated with discordance in the assignment of PICU Up! levels. With respect to HLM, mechanical ventilation (OR = 0.34, 95%CI: 0.19-0.97, P = 0.016), the presence of vascular access lines (OR = 0.43, 95%CI: 0.13-0.69, P = 0.043), and sedation greater than 30 minutes (OR = 0.30, 95%CI: 0.13-0.69, P = 0.005) were associated with reduced odds of discordance when examined independently.
| Characteristics | Discordance in PICU Up! level between expected and RN-report | Discordance in HLM between expected and RN-report | ||
| OR | 95%CI | OR | 95%CI | |
| Age (vs 0-2 years) | ||||
| 3-5 years | 0.91 | 0.24, 3.43 | 0.81 | 0.22, 3.01 |
| 6-12 years | 1.09 | 0.39, 3.04 | 0.71 | 0.25, 1.98 |
| 13-18 years | 0.74 | 0.23, 2.40 | 2.13 | 0.62, 7.29 |
| Gender (vs male) | ||||
| Female | 0.78 | 0.33, 1.72 | 0.72 | 0.33, 1.60 |
| PICU day | 0.73 | 0.52, 1.03 | 0.86 | 0.62, 1.20 |
| ICU reason (vs surgical) | ||||
| Medical | 0.79 | 0.26, 2.40 | 0.81 | 0.26, 2.54 |
| Trauma or other | 1.44 | 0.30, 6.95 | 0.84 | 0.17, 4.13 |
| PRISM score (vs low risk) | ||||
| ≥ Moderate risk | 1.67 | 0.51, 5.40 | 0.46 | 0.14, 1.48 |
| PCPC score (versus normal) | ||||
| Mild disability | 1.31 | 0.41, 4.15 | 2.36 | 0.69, 8.01 |
| Moderate disability | 1.24 | 0.37, 4.19 | 0.80 | 0.25, 2.62 |
| Severe disability | 8.29a | 2.56, 26.91 | 2.49 | 0.87, 7.15 |
| Respiratory support (vs none or nasal cannula) | ||||
| NIPPV or trach collar | 1.06 | 0.31, 3.60 | 2.13 | 0.48, 9.43 |
| Mechanical ventilation | 1.91 | 0.82, 4.47 | 0.34b | 0.14, 0.82 |
| Equipment | ||||
| Lines | 1.34 | 0.62, 2.92 | 0.43b | 0.19, 0.97 |
| Drains and catheters | 0.68 | 0.28, 1.62 | 0.59 | 0.25, 1.41 |
| Gastric/nasogastric tube | 1.97 | 0.85, 4.55 | 1.48 | 0.65, 3.39 |
| No medical equipment | 0.57 | 0.17, 1.91 | 0.31 | 0.093, 1.06 |
| Any sedation > 30 minutes | 0.95 | 0.42, 2.15 | 0.30b | 0.13, 0.69 |
| Any vasoactive within 24 hours | 1.16 | 0.33, 4.07 | 0.66 | 0.19, 2.30 |
| Delirium within 24 hours | 3.24 | 0.46, 22.95 | 0.22 | 0.042, 2.09 |
In multivariate logistic regressions, adjusting for age, PRISM score, and PCPC score, severe disability at baseline increased the odds of a discordant PICU Up! level between RN-report and the expected level compared to patients with good function at baseline (OR = 8.33, 95%CI: 2.50-27.76, P = 0.001) (Figure 2A). Similarly, severe disability at baseline increased the odds of a discordant HLM between RN-report and expected activity (OR = 3.71, 95%CI: 1.15-12.01, P = 0.029) (Figure 2B).
This quality improvement initiative examined the fidelity of RN-reported and EHR-documented mobilizations in a pediatric ICU with a 12-year history of an established early mobility program and identified significant discordances between RN-reported mobility and protocol-derived expectations. We found that nurses were unable to identify the appropriate PICU Up! level for nearly one-third of patients, which often led to patients being mobilized at a lower mobility level compared with protocol recommendations. Furthermore, severe baseline disability (PCPC score) was associated with more than eight times the likelihood of disagreement between RN-reported and expected PICU Up! mobility levels and nearly four times the likelihood of discrepancy between RN-reported and expected HLM achieved, even after adjusting for age and illness severity. Together, these findings suggest that even within a longstanding early mobility program, gaps in fidelity persist, underscoring the need for clearer decision support and targeted implementation strategies.
Pediatric single-center observational studies have demonstrated the efficacy of early mobility in improving clinical outcomes for patients admitted to the ICU[8,9,12,23-25]. Although most prior studies have focused on provider-perceived barriers to early mobility, none have provided granular data on adherence to the newly implemented early mobility protocol. To our knowledge, this is the first study that bridges this gap by reporting on implementation discrepancies in a pediatric early mobility program. We show that the presence of a seasoned program does not guarantee EHR documentation, clinical decision certainty, or implementation fidelity, all of which are prerequisites to the evaluation of program effectiveness[26].
Nurses in pediatric and adult settings have routinely reported barriers to early mobility, such as hemodynamic instability, presence of medical equipment, fear of adverse safety events, staffing concerns, and inadequate training and education as obstacles to mobilization of patients[27-30]. In our cohort, these challenges may be reflected in the substantial number of patients with discordant or uncertain mobility assessments; for example, among the 28 patients with an unknown PICU Up! level, 15 were classified by protocol as level 3, while 16 received an HLM consistent with only level 1 or level 2.
Some deviations from protocol-based mobility targets certainly may reflect appropriate bedside clinical judgment not captured in our protocol criteria. It is possible that nurses were unable to move patients due to competing tasks on the high acuity unit on a particular study day, due to family preferences, and workflow and staffing constraints; these qualitative features were not captured in our dataset[25]. Despite these potential explanations, our findings are consistent with the only adult study evaluating early mobility protocol fidelity, which similarly found that nearly one-quarter of patients were not mobilized as expected[17]. Together, these results suggest that early mobility programs likely experience persistent discordance between protocol expectations and bedside practice, as well as gaps in documentation or classification. This variability has important implications. It contributes to inconsistency in clinical practice for critically ill children and complicates evaluation of early mobility interventions, as program fidelity may vary substantially in routine care. Future studies should report program fidelity between protocol expectations and bedside clinical care to ensure the efficacy of the intervention is being accurately evaluated.
Patients with severe disability at baseline were also more likely to have a discordant PICU Up! level and HLM achieved. For patients with severe disability, the discordance between RN-report and protocol expectations was predominantly driven by nurses being unsure of their mobility level (12/19 patients classified as unknown level). This finding is consistent with decades of research in adult medicine suggesting that patients with chronic conditions, severe disease, and lower baseline functional strength are less likely to be mobilized[27,31]. Similarly, pediatric point-prevalence data demonstrate that children with severe disability are less likely to be mobilized out-of-bed in United States PICUs[14].
Interestingly, in our univariate analysis, mechanical ventilation (OR = 0.34, 95%CI: 0.14-0.82, P = 0.016), vascular access lines (OR = 0.43, 95%CI: 0.19-0.97, P = 0.043), and sedation (OR = 0.30, 95%CI: 0.13-0.69, P = 0.005) significantly decreased the odds of discordance in HLM. An explanation for these findings is that while equipment is clearly accounted for in the PICU Up! mobility guidelines (Figure 1), baseline functional status and the navigation of chronic conditions and disabilities are not. Each patient with severe disability may have individualized baseline functions that vary based on clinical status, location, and availability of mobility aides/home equipment[25]. These factors are not included in most early mobility protocols, including ours. Further, at our center, patients with severe baseline disability, who did not experience a change in their baseline function, did not automatically receive a physical or occupational therapy consult. While nurses remain the key agents of mobilization, not having access to the support of ancillary rehabilitation services for these complex patients may contribute to some of the discordance noted in our study. Taken together, these findings suggest that the discordance may stem less from mechanical barriers and more from systemic ambiguous guidance surrounding children with complex baseline conditions.
Finally, since EHR documentation at the time of mobilization was not standardized in our unit, we relied on RN-reported data as an indicator of bedside care. We observed notable discrepancies between the health record and nurse-reported data. For example, while nurses reported carrying out a median of 6 mobilizations per shift, only a median of 1 was documented in the EHR, suggesting a culture of poor documentation. Similarly, nurse-reported PICU Up! levels and HLM were generally more liberal than values documented in the EHR (Supplementary Table 2). This discordance has important implications for program evaluation. The presence of two non-concordant data sources makes it challenging to determine which more accurately reflects bedside practice and complicates assessment of intervention fidelity. RN-reported data is influenced by recall bias, social desirability bias, and potential overreporting of mobility. Simultaneously, the EHR may underreport these metrics due to inconsistent charting. As a result, neither source can be assumed to perfectly represent clinical practice, which complicates our efforts to assess intervention fidelity. Accurate documentation of mobility is essential for interprofessional communication and continuity of patient care. Incomplete or inconsistent documentation may result in nurses or rehabilitation team members receiving inaccurate information at change of shift, which may contribute further to immobility. Institutions evaluating early mobility programs should prioritize standardized, timely documentation of mobilizations or consider direct observation methods to more objectively assess mobility practices.
Our project is not without limitations. As a single-center quality improvement project at a large, academic center, causal inference is limited, and generalizability requires multi-center evaluation. Several important contextual factors that influence mobility, including nurse-patient staffing ratios, patient census, number of high acuity events per shift, presence of families, and the availability of health professionals such as physical, occupational, and respiratory therapy, were not captured in the original initiative. Further, our primary mobility data relied on nurse self-reporting, which is subject to social desirability bias, recall bias, and over-reporting. Future work by our group includes information on mobility as captured by videos in the patient rooms, which will substantially increase the accuracy of our analysis.
Clinical care involves human factors and situational complications that routinely require departure from protocol. Some deviations from protocol-based mobility targets may reflect appropriate bedside clinical judgment not captured in the criteria. Therefore, a further limitation of this initiative is the absence of qualitative perspectives of nurses on why discordances occurred. Understanding contextual drivers is critical in redesigning protocols to reduce ambiguity and better support mobility champions.
It is also noteworthy that researchers were not blinded to the EHR or RN-reported mobility values while determining the expected protocol-derived mobility goals, which further contributes to bias and possible overreporting of discordance in this study. Our sample size was also small, at 101 total participants. While we employed a penalizing model, namely the Firth Logistic Regression, our confidence intervals were wide, indicating high data variability and imprecision. To increase data transparency, the directionality of discordances noted in this manuscript and the absolute counts in each category, both for PICU Up! level and HLM, are attached in Supplementary Table 2.
The discordances and limitations identified in this study can be understood through the lens of implementation science. Unlike quality improvement, which focuses solely on clinical outcomes, implementation science systematically examines strategies to increase uptake of evidence-backed practices into clinical care[32]. This study demonstrated that adherence to an early mobility protocol was incomplete even in a mature program. A 2021 sustainability analysis of PICU Up! at our institution suggested that resource management (availability of positioning and movement aids), personnel availability, inconsistent sedation, and the heterogeneity of patient characteristics are critical challenges to implemen
Future research should prioritize implementation-focused approaches rather than iterative quality improvement cycles alone. For example, Vasilevskis et al[33] proposed a real-time audit and feedback model with required EHR docu
This quality improvement project is a novel effort to examine the fidelity of protocol adherence and documentation in a mature, long-established PICU early mobility program. Our findings suggest that significant discordance exists between protocol expectations, bedside nursing report, and EHR documentation of mobility. We also show that nurse uncertainty regarding appropriate mobility levels is common and associated with activities corresponding to a lower mobility level than expected. Further, severe baseline disability is independently associated with increased protocol deviations when adjusting for illness severity and age. Future research on early mobility should prioritize evaluation of implementation fidelity, systemic changes to increase access to mobility aides (human and equipment), and optimizing mobility protocols for children with baseline medical complexity.
| 1. | Killien EY, Keller MR, Watson RS, Hartman ME. Epidemiology of Intensive Care Admissions for Children in the US From 2001 to 2019. JAMA Pediatr. 2023;177:506-515. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 74] [Cited by in RCA: 110] [Article Influence: 36.7] [Reference Citation Analysis (0)] |
| 2. | Manning JC, Pinto NP, Rennick JE, Colville G, Curley MAQ. Conceptualizing Post Intensive Care Syndrome in Children-The PICS-p Framework. Pediatr Crit Care Med. 2018;19:298-300. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 184] [Cited by in RCA: 382] [Article Influence: 47.8] [Reference Citation Analysis (0)] |
| 3. | Woodruff AG, Choong K. Long-Term Outcomes and the Post-Intensive Care Syndrome in Critically Ill Children: A North American Perspective. Children (Basel). 2021;8:254. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 9] [Cited by in RCA: 48] [Article Influence: 9.6] [Reference Citation Analysis (2)] |
| 4. | Carlton EF, Gebremariam A, Maddux AB, McNamara N, Barbaro RP, Cornell TT, Iwashyna TJ, Prosser LA, Zimmerman J, Weiss S, Prescott HC. New and Progressive Medical Conditions After Pediatric Sepsis Hospitalization Requiring Critical Care. JAMA Pediatr. 2022;176:e223554. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1] [Cited by in RCA: 20] [Article Influence: 5.0] [Reference Citation Analysis (0)] |
| 5. | Senna S, Ong C, Wong JJ, Allen JC Jr, Sultana R, Lee JH. Prediction of Acquired Morbidity Using Illness Severity Indices in Pediatric Intensive Care Patients. Pediatr Crit Care Med. 2020;21:e972-e980. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 8] [Cited by in RCA: 16] [Article Influence: 2.7] [Reference Citation Analysis (0)] |
| 6. | Schweickert WD, Pohlman MC, Pohlman AS, Nigos C, Pawlik AJ, Esbrook CL, Spears L, Miller M, Franczyk M, Deprizio D, Schmidt GA, Bowman A, Barr R, McCallister KE, Hall JB, Kress JP. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373:1874-1882. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 2647] [Cited by in RCA: 2217] [Article Influence: 130.4] [Reference Citation Analysis (0)] |
| 7. | Wieczorek B, Burke C, Al-Harbi A, Kudchadkar SR. Early mobilization in the pediatric intensive care unit: a systematic review. J Pediatr Intensive Care. 2015;2015:129-170. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 24] [Cited by in RCA: 39] [Article Influence: 3.5] [Reference Citation Analysis (0)] |
| 8. | Betters KA, Hebbar KB, Farthing D, Griego B, Easley T, Turman H, Perrino L, Sparacino S, deAlmeida ML. Development and implementation of an early mobility program for mechanically ventilated pediatric patients. J Crit Care. 2017;41:303-308. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 27] [Cited by in RCA: 38] [Article Influence: 4.2] [Reference Citation Analysis (0)] |
| 9. | Choong K, Awladthani S, Khawaji A, Clark H, Borhan A, Cheng J, Laskey S, Neu C, Sarti A, Thabane L, Timmons BW, Zheng K, Al-Harbi S; Canadian Critical Care Trials Group. Early Exercise in Critically Ill Youth and Children, a Preliminary Evaluation: The wEECYCLE Pilot Trial. Pediatr Crit Care Med. 2017;18:e546-e554. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 28] [Cited by in RCA: 44] [Article Influence: 4.9] [Reference Citation Analysis (0)] |
| 10. | Herbsman JM, D'Agati M, Klein D, O'Donnell S, Corcoran JR, Folks TD, Al-Qaqaa YM. Early Mobilization in the Pediatric Intensive Care Unit: A Quality Improvement Initiative. Pediatr Qual Saf. 2020;5:e256. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 7] [Cited by in RCA: 24] [Article Influence: 4.0] [Reference Citation Analysis (0)] |
| 11. | Wieczorek B, Ascenzi J, Kim Y, Lenker H, Potter C, Shata NJ, Mitchell L, Haut C, Berkowitz I, Pidcock F, Hoch J, Malamed C, Kravitz T, Kudchadkar SR. PICU Up!: Impact of a Quality Improvement Intervention to Promote Early Mobilization in Critically Ill Children. Pediatr Crit Care Med. 2016;17:e559-e566. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 181] [Cited by in RCA: 162] [Article Influence: 16.2] [Reference Citation Analysis (0)] |
| 12. | Adel TZD, van Dijk M, de Heer M, Hoekstra S, Steenhorst J, van Rosmalen J, Verbruggen S, Toussaint-Duyster L, Ista E. Quality improvement intervention to stimulate early mobilisation of critically ill children. Nurs Crit Care. 2023;28:545-553. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1] [Cited by in RCA: 12] [Article Influence: 3.0] [Reference Citation Analysis (0)] |
| 13. | Colwell BRL, Williams CN, Kelly SP, Ibsen LM. Mobilization Therapy in the Pediatric Intensive Care Unit: A Multidisciplinary Quality Improvement Initiative. Am J Crit Care. 2018;27:194-203. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 17] [Cited by in RCA: 23] [Article Influence: 2.9] [Reference Citation Analysis (0)] |
| 14. | Kudchadkar SR, Nelliot A, Awojoodu R, Vaidya D, Traube C, Walker T, Needham DM; Prevalence of Acute Rehabilitation for Kids in the PICU (PARK-PICU) Investigators and the Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network. Physical Rehabilitation in Critically Ill Children: A Multicenter Point Prevalence Study in the United States. Crit Care Med. 2020;48:634-644. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 84] [Cited by in RCA: 84] [Article Influence: 14.0] [Reference Citation Analysis (1)] |
| 15. | Noone CE, Franck LS, Staveski SL, Rehm RS. Barriers and facilitators to early mobilisation programmes in the paediatric intensive care unit: A scoping literature review. Nurs Crit Care. 2023;28:526-544. [RCA] [DOI] [Full Text] [Cited by in RCA: 7] [Reference Citation Analysis (0)] |
| 16. | Costa DK, White MR, Ginier E, Manojlovich M, Govindan S, Iwashyna TJ, Sales AE. Identifying Barriers to Delivering the Awakening and Breathing Coordination, Delirium, and Early Exercise/Mobility Bundle to Minimize Adverse Outcomes for Mechanically Ventilated Patients: A Systematic Review. Chest. 2017;152:304-311. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 72] [Cited by in RCA: 134] [Article Influence: 14.9] [Reference Citation Analysis (3)] |
| 17. | Perelló P, Gómez J, Mariné J, Cabas MT, Arasa A, Ramos Z, Moya D, Reynals I, Bodí M, Magret M. Analysis of adherence to an early mobilization protocol in an intensive care unit: Data collected prospectively over a period of three years by the clinical information system. Med Intensiva (Engl Ed). 2023;47:203-211. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 18. | LaRosa JM, Lenker H, Mennie C, Morgenstern S, Furniturewala S, Hwang L, Rao NN, Hajnik K, Brown KM, Shilkofski N, Kudchadkar SR. Increasing Nurse-Led Mobilization of Critically Ill Children Through In Situ Simulation: A Quality Improvement Initiative. Pediatr Crit Care Med. 2026;27:62-71. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1] [Cited by in RCA: 2] [Article Influence: 2.0] [Reference Citation Analysis (0)] |
| 19. | Fiser DH. Assessing the outcome of pediatric intensive care. J Pediatr. 1992;121:68-74. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 650] [Cited by in RCA: 774] [Article Influence: 22.8] [Reference Citation Analysis (0)] |
| 20. | Pollack MM, Patel KM, Ruttimann UE. PRISM III: an updated Pediatric Risk of Mortality score. Crit Care Med. 1996;24:743-752. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1185] [Cited by in RCA: 1370] [Article Influence: 45.7] [Reference Citation Analysis (1)] |
| 21. | Viera AJ, Garrett JM. Understanding interobserver agreement: the kappa statistic. Fam Med. 2005;37:360-363. [PubMed] |
| 22. | Li M, Gao Q, Yu T. Kappa statistic considerations in evaluating inter-rater reliability between two raters: which, when and context matters. BMC Cancer. 2023;23:799. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 89] [Reference Citation Analysis (0)] |
| 23. | Tsuboi N, Hiratsuka M, Kaneko S, Nishimura N, Nakagawa S, Kasahara M, Kamikubo T. Benefits of Early Mobilization After Pediatric Liver Transplantation. Pediatr Crit Care Med. 2019;20:e91-e97. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 17] [Cited by in RCA: 27] [Article Influence: 3.9] [Reference Citation Analysis (0)] |
| 24. | Simone S, Edwards S, Lardieri A, Walker LK, Graciano AL, Kishk OA, Custer JW. Implementation of an ICU Bundle: An Interprofessional Quality Improvement Project to Enhance Delirium Management and Monitor Delirium Prevalence in a Single PICU. Pediatr Crit Care Med. 2017;18:531-540. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 82] [Cited by in RCA: 107] [Article Influence: 11.9] [Reference Citation Analysis (0)] |
| 25. | Patel RV, Redivo J, Nelliot A, Eakin MN, Wieczorek B, Quinn J, Gurses AP, Balas MC, Needham DM, Kudchadkar SR. Early Mobilization in a PICU: A Qualitative Sustainability Analysis of PICU Up! Pediatr Crit Care Med. 2021;22:e233-e242. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 15] [Cited by in RCA: 29] [Article Influence: 5.8] [Reference Citation Analysis (0)] |
| 26. | Carroll C, Patterson M, Wood S, Booth A, Rick J, Balain S. A conceptual framework for implementation fidelity. Implement Sci. 2007;2:40. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1069] [Cited by in RCA: 1365] [Article Influence: 71.8] [Reference Citation Analysis (0)] |
| 27. | Hanna ES, Zhao S, Shannon CN, Betters KA. Changes in Provider Perceptions Regarding Early Mobility in the PICU. Pediatr Crit Care Med. 2020;21:e30-e38. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 10] [Cited by in RCA: 18] [Article Influence: 3.0] [Reference Citation Analysis (0)] |
| 28. | Hoyer EH, Brotman DJ, Chan KS, Needham DM. Barriers to early mobility of hospitalized general medicine patients: survey development and results. Am J Phys Med Rehabil. 2015;94:304-312. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 82] [Cited by in RCA: 117] [Article Influence: 10.6] [Reference Citation Analysis (0)] |
| 29. | Kim C, Kim S, Yang J, Choi M. Nurses' perceived barriers and educational needs for early mobilisation of critical ill patients. Aust Crit Care. 2019;32:451-457. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 8] [Cited by in RCA: 22] [Article Influence: 3.1] [Reference Citation Analysis (0)] |
| 30. | Noone CE, Franck LS, Staveski SL, Rehm RS. Overcoming patient safety concerns and integrating early mobility into pediatric intensive care unit nursing practice. J Pediatr Nurs. 2023;73:e107-e115. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 3] [Reference Citation Analysis (0)] |
| 31. | Zhang H, Bai S, Sheng Y, Dai Z, Li Z, Luo H, Zhang Q, Lin FF. Quantifying Patient-Level Factors Associated With Mobilisation in Intensive Care: A Prospective Study. J Clin Nurs. 2026;35:813-821. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 32. | Yang Y, Geva A, Madden K, Mehta NM. Implementation Science in Pediatric Critical Care - Sedation and Analgesia Practices as a Case Study. Front Pediatr. 2022;10:864029. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1] [Cited by in RCA: 9] [Article Influence: 2.3] [Reference Citation Analysis (0)] |
| 33. | Vasilevskis EE, Kim J, Wichman C, Horner R, Campbell JR, Geary CR, Birge J, Hetland B, Blanchard A, Circo K, Wagner-Connolly A, Miller J, Exline MC, Gerlach AT, Krupp A, Blum J, Balas MC; BEST-ICU Team. Behavioral Economic and Staffing Strategies To Increase Adoption of the ABCDEF Bundle in the Intensive Care Unit (BEST ICU): study protocol for a multi-site stepped-wedge cluster randomized controlled trial in the USA. Trials. 2025;27:42. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |