AS THE POPULATION AGES and mortality from critical

536 ORIGINAL ARTICLE Early Physical Medicine and Rehabilitation for Patients With Acute Respiratory Failure: A Quality Improvement Project Dale M. Needham, MD, PhD, Radha Korupolu, MBBS, MS, Jennifer M. Zanni, PT, MSPT, Pranoti Pradhan, MBBS, MPH, Elizabeth Colantuoni, PhD, Jeffrey B. Palmer, MD, Roy G. Brower, MD, Eddy Fan, MD ABSTRACT. Needham DM, Korupolu R, Zanni JM, Pradhan P, Colantuoni E, Palmer JB, Brower RG, Fan E. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil 2010;91:536-42. Objectives: To (1) reduce deep sedation and delirium to permit mobilization, (2) increase the frequency of rehabilitation consultations and treatments to improve patients functional mobility, and (3) evaluate effects on length of stay. Design: Seven-month prospective before/after quality improvement project. Setting: Sixteen-bed medical intensive care unit (MICU) in academic hospital. Participants: 57 patients mechanically ventilated 4 days or longer. Intervention: A multidisciplinary team focused on reducing heavy sedation and increasing MICU staffing to include fulltime physical and occupational therapists with new consultation guidelines. Main Outcome Measures: Sedation and delirium status, rehabilitation treatments, functional mobility. Results: Compared with before the quality improvement project, benzodiazepine use decreased markedly (proportion of MICU days that patients received benzodiazepines [50% vs 25%, P.002]), with lower median daily sedative doses (47 vs 15mg midazolam equivalents [P.09] and 71 vs 24 mg morphine equivalents [P.01]). Patients had improved sedation and delirium status (MICU days alert [30% vs 67%, P.001] and not delirious [21% vs 53%, P.003]). There were a greater median number of rehabilitation treatments per patient (1 vs 7, P.001) with a higher level of functional mobility (treatments involving sitting or greater mobility, 56% vs 78%, P.03). Hospital administrative data demonstrated that across all MICU patients, there was a decrease in intensive care unit and hospital length of stay by 2.1 (95% confidence interval: 0.4 3.8) and 3.1 From the Department of Physical Medicine and Rehabilitation (Needham, Zanni, Palmer), the OACIS Group, Division of Pulmonary and Critical Care Medicine (Needham, Korupolu, Pradhan, Brower, Fan), and the Department of Anesthesiology and Critical Care Medicine (Colantuoni), School of Medicine, and the Department of Biostatistics, Bloomberg School of Public Health, (Colantuoni), Johns Hopkins University, Baltimore, MD. Presented to the American Academy of Physical Medicine and Rehabilitation, October 25, 2009, Austin, TX. Supported by the Department of Physical Medicine and Rehabilitation and the Division of Pulmonary and Critical Care Medicine, Johns Hopkins University. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Correspondence to Dale M. Needham, MD, PhD, 1830 E Monument St, 5th Fl, Baltimore, MD 21205, e-mail: dale.needham@jhmi.edu. Reprints are not available from the author. 0003-9993/10/9104-00878$36.00/0 doi:10.1016/j.apmr.2010.01.002 (0.3 5.9) days, respectively, and a 20% increase in MICU admissions compared with the same period in the prior year. Conclusions: Using a quality improvement process, intensive care unit delirium, physical rehabilitation, and functional mobility were markedly improved and associated with decreased length of stay. Key Words: Critical care; Early ambulation; Muscle weakness; Rehabilitation; Respiration, artificial. 2010 by the American Congress of Rehabilitation Medicine AS THE POPULATION AGES and mortality from critical illness declines, the number of ICU survivors is growing. 1,2 These survivors commonly experience neuromuscular weakness that may be severe and prolonged. 1-3 Particularly in mechanically ventilated patients, heavy sedation and bed rest are common in the ICU. 4-6 Immobility plays an important role in the development of neuromuscular weakness, 7,8 which is associated with impairment in ICU survivors physical function, quality of life, and return to work. 3,9 Physical inactivity also contributes to the development of atelectasis, insulin resistance, and joint contractures. 10,11 Mobilizing mechanically ventilated patients in the ICU has a historical precedent and has been demonstrated as feasible, safe, and beneficial in improving physical function. 12-14 However, early mobilization is not practiced on a widespread basis in ICUs. 15 Neuromuscular complications after critical illness and early mobilization of mechanically ventilated patients became areas of interest at our institution. Our university is the lead institution for a multisite prospective cohort study evaluating the long-term physical and mental health outcomes of patients who survived acute lung injury/acute respiratory distress syndrome. 16 Experience from follow-up of these research participants increased awareness of the prolonged neuromuscular complications faced by patients discharged from our MICU. Moreover, analysis of preliminary data from this study demonstrated that only 24% of patients ever received consultation for PT and/or OT in our MICU, which was almost 50% lower than for similar patients at 2 other academic hospitals in the same city. 17 These data also demonstrated a higher prevalence of deep sedation in our MICU patients (58% vs 27% of ICU CI ICU LOS MICU OT PM&R PT QI List of Abbreviations confidence interval intensive care unit length of stay medical intensive care unit occupational therapy physical medicine and rehabilitation physical therapy quality improvement

EARLY PHYSICAL MEDICINE AND REHABILITATION FOR ACUTE RESPIRATORY FAILURE, Needham 537 patient days) and a low proportion ( 15%) of ICU days in which patients were not deeply sedated or delirious. Additional observational work in our MICU further motivated the need for quality improvement through confirming that heavy sedation represented an important barrier to implementing early PM&R and that our MICU survivors experienced important impairments in strength, range of motion, and physical function at hospital discharge. 18 Based on this experience, we wanted to improve PM&R services in the MICU. Focusing on a population of MICU patients mechanically ventilated for 4 days or longer, the goals of this before/after QI project were to (1) optimize patient sedation and analgesia practices, reducing deep sedation and delirium to permit PM&R therapies while maintaining patient comfort; and (2) increase the frequency of PM&R-related consultation and treatment to improve patients functional mobility in the MICU. Using hospital administrative data, we aimed to evaluate the effect of the QI project, across all MICU patients, on the number of PT and OT consultations/treatments and length of stay, in comparison with the prior year. METHODS Overview of Project Design and Timing This multifaceted QI project was conducted using a structured QI framework and evaluated using a before/after design. The initial phases of the QI project (ie, the engage and educate processes, as described in the Quality Improvement Process section) started in spring 2006 with increasing intensity until the 4-month execution phase (May to August 2007), during which early PM&R was implemented. For purposes of the before/after comparison, this execution phase is referred to as the QI period and is compared with the immediately preceding 3-month pre-qi period (February to April 2007). During the entire 7-month combined pre-qi and QI periods, prospective collection of relevant data occurred for the target patient population. To further evaluate the overall effects of the QI project on all MICU patients, data regarding the number of PT and OT consultations/treatments and LOS were obtained from hospital administrative data to compare the QI period with the same 4-month period in the prior year (ie, May to August 2006). The prior year was used in this latter comparison in order to control for known seasonal effects on the number of MICU admissions and LOS. Setting The MICU at our hospital has 16 beds and is staffed with attending, fellow, and resident physicians and registered nurses (staff-to-patient ratio1:2) and respiratory therapists (staff-topatient ratio 1:8). Neurology consultation and PT and OT are available when ordered by an MICU physician. Physiatry consultation did not occur while patients were in the MICU. In the MICU, bed rest was the prescribed activity level in standard admission orders, and there were no MICU guidelines for consultation or treatment by a PT or OT. Routine nursing care included repositioning patients in bed every 2 hours and the use of standardized pain and sedation scales, with a nurse-titrated sedation protocol and a daily reduction in sedation infusions. 19 Standardized assessments for delirium in the MICU were not part of routine nursing care. Patients In both the pre-qi and QI periods, we targeted prospective data collection regarding patients baseline status and outcomes for the patients who we felt would derive the greatest benefit from increased PM&R therapy. Specifically, we focused on MICU patients who were cognitively intact without neuromuscular disease prior to MICU admission and required mechanical ventilation for 4 days or longer. Quality Improvement Process We used a structured QI model, 20 which included the following components: (1) understanding the problem within the larger healthcare system, (2) creating a multidisciplinary improvement team, (3) enlisting all stakeholders to identify barriers to change and appropriate solutions, and (4) creating a change in practice through a 4 Es approach: engage, educate, execute, and evaluate. Many meetings, led by the project leader (DMN), were required to reach the full complement of 66 MICU nurses, 45 respiratory therapists, 13 attending physicians, and 12 pulmonary and critical care fellows who work in the MICU. Moreover, within the Department of PM&R, meetings were held with the director (JBP), physicians, and PT and OT supervisors and staff. Similar meetings were held with the leadership and resident physicians within the Department of Neurology and its neuromuscular subspecialty physician group. These meetings aimed at presenting the problem (as previously outlined) and identifying barriers and solutions for reaching the project goals. A multidisciplinary QI team with representatives from each relevant clinician group in the MICU and PM&R was created and met on a weekly basis to plan, execute, and evaluate the QI project. The process for improving practice was based on a 4 Es model (engage, educate, execute, and evaluate). 20 First, in addition to the multidisciplinary meetings previously described, further steps were taken to engage all relevant stakeholders in the QI process, including (1) providing information about the project in separate MICU and hospital-wide newsletters, (2) creating informational posters, (3) conducting didactic conferences and presentations, and (4) arranging visits by patients to share their stories of neuromuscular weakness after MICU discharge. Furthermore, patients who participated in early PM&R therapy returned to the MICU to provide positive feedback to clinicians about their MICU experiences and subsequent recovery process. Patient interviews and visits reinforced the perceived benefits of decreased sedation and increased PM&R therapy and activity level, without increased patient anxiety, distress, or pain (videos of patient interviews available at www.hopkinsmedicine.org/oacis). Second, education was provided via meetings, presentations, and communications that summarized research publications on long-term neuromuscular complications after critical illness and benefits of early PM&R activities in the ICU. A published expert in this field was invited for a 2-day visit to our institution to give presentations and meet with all stakeholder groups. In addition, a PT leader (JMZ), the MICU physician director (RGB), and a senior MICU nurse visited an ICU that was highly successful with early mobilization and shared the learning from this site visit with their clinical colleagues at our institution. Interdisciplinary education and training was held among the nurses, PTs, OTs, and respiratory therapists to obtain specific skills relevant to rehabilitation of mechanically ventilated patients. Finally, 16 educational sessions were held to inform all MICU nurses regarding sedation-related issues within the QI project. Third, execution of the project during the 4-month QI period involved the following steps:

538 EARLY PHYSICAL MEDICINE AND REHABILITATION FOR ACUTE RESPIRATORY FAILURE, Needham 1. Modifying the standardized MICU admission orders to change the default activity level from bed rest to as tolerated. 2. Encouraging a change in sedation practice from use of continuous intravenous infusions of benzodiazepines and narcotics to as needed bolus doses. 21,22 3. Establishing and disseminating simple guidelines for PT and OT consultation. 4. Developing safety-related guidelines (developed from the existing literature 12 ) regarding when patients were considered eligible for PM&R-related consultation, which included (a) were not comatose (ie, Richmond Agitation-Sedation Score 23 3), (b) required only moderate ventilatory support (ie, positive end-expiratory pressure 10cmH 2 O and fraction of inspired oxygen 0.6), and (c) had no increase in the dose of any vasopressor infusion (used for management of hypotension/shock) for at least 2 hours. 5. Changing staffing to include a full-time PT and OT and a part-time rehabilitation assistant. 6. Consulting a physiatrist for MICU patients receiving rehabilitation therapy. 7. Increasing consultations to neurologists for MICU patients with muscle weakness that is severe or prolonged. Fourth, evaluation of the project occurred on an ongoing basis during the QI period via weekly meetings of the multidisciplinary QI project team to discuss progress, barriers, and solutions. Data Collection For all patients included during the 3-month pre-qi period 18 and the 4-month QI period, data from paper and electronic medical records were abstracted, and relevant evaluations were completed as described in the following paragraphs. Baseline data. Patient baseline data including demographics, comorbidities (including the Charlson Index 24 ), and severity of illness at ICU admission were obtained from the medical record. Sedation, delirium, and medication outcomes. For included patients, the following data were collected on a daily basis while in the MICU: (1) benzodiazepine and narcotic drug doses received (converted to midazolam- and morphine-equivalent doses, respectively, using standard conversion factors 25,26 ), (2) sedation and delirium status (evaluated using the validated Richmond Agitation-Sedation Scale 23 and Confusion Assessment Method for the ICU 27 instruments, respectively), and (3) patient pain status (based on MICU nurses routine clinical assessments using a standard 0 10 scale, with a higher number representing greater pain). PM&R outcomes. The number of PM&R-related consultations and treatments occurring while each patient was in the MICU was collected. In addition, daily functional mobility activities conducted by PT and OT were recorded by the therapist using standard categories from prior related research. 12 Unexpected events occurring during PT and OT (defined as cardiopulmonary arrest, loss of consciousness, fall, removal of any medical device, or oxygen desaturation 85% for 3 minutes) were prospectively evaluated with each treatment. Hospital administrative data for all MICU patients. In order to evaluate any overall impact of the QI project across all MICU patients, hospital administrative data were evaluated. Specifically, the number of PT and OT consultations and treatments and the number of admissions and LOS for all patients receiving care in the MICU during the 4-month QI period and the same period in the prior year were obtained from by the Departments of PM&R and Medicine, respectively. Statistical Analysis Descriptive statistics including proportions (for binary and categorical data) and medians with interquartile range (for continuous data) were used to summarize individual patientlevel data and the data collected on a daily basis during patients MICU stay. Fisher exact and Wilcoxon rank-sum tests were used to compare the patient-level demographic data and MICU outcomes (including receipt of any benzodiazepine, narcotic, or physical and/or occupational therapy) in the pre-qi versus QI periods. Daily use and dose of benzodiazepine and narcotics, daily sedation and delirium status, and daily functional mobility measures were compared across the pre-qi and QI periods using linear, logistic, and multinomial regression models with robust variance estimates to account for the correlation of repeated daily measures from the same person during their MICU stay. 28 For linear regression analyses of midazolam- and morphine-equivalent drug doses, data were log-transformed. T tests were used to evaluate the difference in average ICU and hospital LOS comparing the pre-qi and QI periods. All analyses were performed using Stata 10.0 software. a A 2-sided P value less than.05 was used to determine statistical significance. A detailed description of the proposed project was provided to the institutional review board Chair. On review of the project, it was considered to be quality improvement in nature and thus did not require institutional review board approval. This QI project was reported in accordance with the Standards for Quality Improvement Reporting Excellence guidelines. 29 Table 1: Baseline Characteristics of Patients Eligible for PM&R Therapy Baseline Characteristics Pre-QI Period (n 27 Patients) QI Period (n 30 Patients) P* Demographics Age, median (IQR) years 50 (43 59) 53 (43 69).35 Female 19 (70) 21 (70).99 Race White 13 (48) 13 (43).79 Black 14 (52) 17 (57) Baseline medical data Charlson Comorbidity Index, median (IQR) 2 (1 3) 4 (1 6).08 Chronic obstructive pulmonary disease 11 (41) 16 (53).43 End-stage renal disease on dialysis 1 (4) 6 (20).11 MICU data APACHE II score, median (IQR) 26 (21 29) 27 (21 32).43 ALI/ARDS 7 (26) 10 (33).58 NOTE. Values are n (%) or as otherwise indicated. Abbreviations: ALI/ARDS, acute lung injury/acute respiratory distress syndrome; APACHE, Acute Physiology and Chronic Health Evaluation; IQR, interquartile range. *Fisher exact and Wilcoxon rank-sum tests were used to compare patient characteristics across the pre-qi and QI period. The Charlson Comorbidity Index is a measure of comorbid disease calculated as the weighted sum of 19 comorbid conditions with a score ranging from 0 to 37; higher scores signify greater comorbidity and a higher risk of short-term mortality. 24

EARLY PHYSICAL MEDICINE AND REHABILITATION FOR ACUTE RESPIRATORY FAILURE, Needham 539 Outcome Measure Table 2: Sedation, Delirium, and Medication Outcomes Pre-QI Period (n 27 Patients with 312 MICU Patient Days) QI Period (n 30 Patients With 482 MICU Patient Days) P* Benzodiazepines Patients ever receiving benzodiazepines 26 (96) 22 (73).030 MICU days with any benzodiazepine use 150 (50) 118 (26).002 Daily midazolam-equivalent dose, units (median [IQR] units) 47 (21 114) 15 (3 59).090 Narcotics Patients ever receiving narcotics 26 (96) 23 (77).050 MICU days with any narcotic use 188 (62) 299 (66).650 Daily morphine-equivalent dose (median [IQR] units) 71 (30 180) 24 (3 120).010 Pain Daily scores (range, 0 10) (mean SD) 0.6 1.9 0.6 1.7.790 Sedation status (daily RASS 23 ) of MICU days Deeply sedated (RASS 4 to 5) 129 (43) 86 (18).001 Moderately sedated (RASS 2 to 3) 72 (24) 65 (14) Alert (RASS 1 to 1) 88 (30) 311 (67) Agitated (RASS 2 to 4) 8 (3) 6 (1) Delirium status (daily CAM-ICU 27 ) of MICU days Delirious 107 (36) 125 (28).003 Not delirious 61 (21) 243 (53) Unable to assess because of deep sedation 129 (43) 86 (19) NOTE. Values are n (%) or as otherwise indicated. Abbreviations: CAM-ICU, Confusion Assessment Method for the ICU; IQR, interquartile range; RASS, Richmond Agitation-Sedation Scale. *Fisher exact test was used to compare the proportion of patients during the pre-qi and QI periods with any benzodiazepines or narcotics. For daily MICU measures, the P values were estimated from linear and multinomial regression models adjusting for within-subject correlation using a robust variance estimate. 28 For linear regression analyses of midazolam- and morphine-equivalent dose, data were log-transformed. There were missing data for daily benzodiazepine and narcotic drug dose for 9 and 28 MICU patient days during the pre-qi and QI period, respectively. There were no daily nursing assessments of pain status, because patients were deeply sedated, for 109 and 91 MICU patient days during the pre-qi and QI period, respectively. There were no daily sedation assessments, because patients were unavailable for assessment for 15 and 14 MICU patient days during the pre-qi and QI period, respectively. There were no daily delirium assessments, because of patient refusal or being unavailable for assessment, for 15 and 28 MICU patient days during the pre-qi and QI period, respectively. RESULTS All eligible MICU patients during the pre-qi and QI periods were included in the project, representing a total of 27 and 30 patients requiring 312 and 482 MICU patient days, respectively. These patients represented approximately 10% of all MICU admissions during each of the 2 time periods. Compared with the immediately prior pre-qi period, patients in the QI period tended to be slightly older with greater comorbidities at baseline and greater severity of illness in the MICU (table 1). Table 3: PM&R Outcomes Outcome Measure Pre-QI Period QI Period P* Patient data Patients (n 27) Patients (n 30) Received physical and/or occupational therapy in MICU 19 (70) 28 (93).040 Number of treatments per patient (median [IQR]) 1 (0 3) 7 (3 15).001 Number of treatments per day (mean SD) 0.33 0.38 0.83 0.45.001 Consultations Physical therapy 16 (59) 28 (93).004 Occupational therapy 20 (74) 27 (90).170 Physiatry 0 (0) 26 (87).001 Neurology 1 (4) 7 (23).050 Functional mobility during a PT or OT treatment Treatments (n 50) Treatments (n 294) Supine to sit 19 (38) 212 (72).003 Sitting at the edge of the bed 27 (54) 225 (77).020 Transfer from bed to chair 3 (6) 113 (38).005 Transfer from sit to stand 12 (24) 145 (49).050 Walking 2 (4) 39 (13).240 NOTE. Values are n (%) or as otherwise indicated. Abbreviation: IQR, interquartile range. *Fisher exact test was used to compare the proportion of patients that received any physical and/or occupational therapy and by type of therapy across the pre-qi and QI periods. For measures recorded on multiple days in the MICU, P values were estimated using logistic regression models adjusting for within-subject correlation using a robust variance estimate.

540 EARLY PHYSICAL MEDICINE AND REHABILITATION FOR ACUTE RESPIRATORY FAILURE, Needham Sedation, Delirium, and Medication Outcomes With respect to the first objective of the QI project, in comparison with the pre-qi period, we found that a lower proportion of MICU patients received benzodiazepines (96% vs 73%, P.03) and narcotics (96% vs 77%, P.05). There was a large decrease in the proportion of MICU days in which patients received benzodiazepines (50% vs 26%, P.002), but not narcotics (62% vs 66%, P.65) with lower median doses given (47 vs 15mg of midazolam equivalents [P.09], 71 vs 24mg of morphine equivalents [P.01]) (table 2). Moreover, we found that patients were more frequently alert (29% vs 66% of MICU days, P.001) and not delirious (21% vs 53%, P.003). Patients in both periods similarly had very low pain scores, based on routine nursing assessments using a 0 to 10 scale (0.6 vs 0.6, P.79). PM&R Outcomes With respect to the second objective of this project, during the QI period, important barriers to rehabilitation therapy were surmounted. There was a substantial increase in the proportion of patients who received PT and/or OT therapy in the MICU (70% vs 93%, P.04) and PM&R-related consultations (table 3). These improvements led to a substantial decrease in the proportion of MICU days in which eligible patients failed to receive any Fig 1. A 56-year old man during his fourth day in the medical intensive care unit is ambulating, while receiving mechanical ventilation via an oral endotracheal tube, with the assistance of a physical therapist, respiratory therapist, and an ICU nurse. The associated equipment includes a portable ventilator with attached oxygen tanks, a portable cardiac monitor, a wheeled pole with intravenous infusion pumps, and a wheeled walker. A wheelchair (not seen) is being pushed behind the patient by a technician. Table 4: Hospital Administrative Data for All MICU Patients: Treatments and LOS Outcome Measure Control Period* (May August 2006) QI Period (May August 2007) Relative Change (%) P Physical and occupational therapy Total consultations 215 548 1154.040 Total treatments 210 810 1286.001 Data for MICU patients Number of admissions 262 314 120 MICU average LOS, d 7.0 4.9 230.020 Hospital average LOS, d 17.2 14.1 218.030 In-hospital mortality (%) 23.3 21.0 210.550 Abbreviations: 1, increase; 2, decrease. *The same 4-month period from the prior year was used as a control period for comparison with the QI period when using hospital administrative data to evaluate the effect of the QI project on all MICU patients. A prior year comparison period was used, rather than the months immediately preceding the QI period, in order to control for known seasonal effects in the number of MICU admissions and length of stay. therapy from a PT and/or OT (41% vs 7%, P.004). Moreover, there were more PT and OT treatments per eligible patient (median 1 vs 7, P.001) and per eligible MICU day (mean.33 vs.83, P.001), with a greater proportion of these treatments (56% vs 78%, P.03) having a functional mobility level of sitting or greater (see table 3; fig 1). In the QI period, the only prospectively defined unexpected events during PM&R therapy were 4 instances in which a rectal or feeding tube was displaced or removed, without any consequential medical complications versus no unexpected events in the pre-qi period (P.99). These specific events were not unique to PM&R therapy because they had also occurred in the context of routine nursing care. Hospital Administrative Data for All MICU Patients Hospital administrative data allowed additional analyses to be performed for all MICU patients during the QI period rather than only the subgroup of patients mechanically ventilated 4 days or longer who were the focus of the results described in the prior paragraphs. For these analyses, all MICU patients from the same 4-month period in the prior year (n 262) were compared with patients in the 4-month QI period (n 314). Comparing these two 4-month time periods, there were significant 2- to 4-fold increases in the combined number of PT and OT consultations and treatments, with an almost 5-fold increase (.11 vs.53) in the average number of treatments per MICU patient day (table 4). Moreover, there was a decrease in the average MICU LOS by 2.1 days (95% CI, 0.4 3.8d) and in the average hospital LOS by 3.1 days (95% CI, 0.3 5.9d), with a 20% increase in MICU admissions and no significant change in in-hospital mortality for MICU patients. DISCUSSION Through a structured model for QI, we learned that deep sedation was generally not necessary for patients comfort and tolerance of mechanical ventilation. Moreover, with a change in sedation practice, ICU delirium was substantially lower and early PM&R was feasible and safe, with increased functional mobility in the MICU and substantially decreased LOS. To our knowledge, given the relatively recent onset of interest in early PM&R in ICUs in the United States, there are no

EARLY PHYSICAL MEDICINE AND REHABILITATION FOR ACUTE RESPIRATORY FAILURE, Needham 541 prior published QI reports in this area. However, as the foundation of evidence-based medicine increases, both small- and large-scale QI initiatives, and related QI methodology, are gaining prominence within critical care medicine. 20,30-32 Our QI project is set within the context of a growing interest in early PM&R in the ICU. 33-35 Historically, early ambulation of hospitalized patients appears to have gained interest in the 1940s 36,37 and occurred, at least in some ICUs, during the first few decades after the inception of ICUs. 38,39 However, research evidence supporting the benefits of early mobilization of critically ill patients has only been published more recently and includes an initial landmark study of 103 consecutive patients 12 followed by a subsequent larger, nonrandomized controlled trial 13 and then a 2-site randomized controlled trial. 14 In these studies, increased mobilization, improved physical function, and decreased ICU resource utilization have been clearly demonstrated and were the foundation for our QI project. 13,14,40 Moreover, evidence supporting the short- and long-term benefits of reducing deep sedation, including decreased delirium and ICU resource utilization, has also evolved over the past 30 years with the introduction and validation of sedation scales, goal-directed sedation, interruption of continuous sedative infusions, use of bolus (rather than infusion) delivery of sedatives, and novel sedative agents. 19,21-23,41-44 This QI project applied evidence from this body of literature and demonstrated that within a relatively short time period a large change in routine clinical practice could occur and achieve benefits similar to those demonstrated in prior research studies. As part of continuous QI efforts, several steps have been taken to achieve further advances regarding early PM&R in the MICU at our hospital. Given the benefits demonstrated from this project, the hospital funded a new Critical Care Physical Medicine and Rehabilitation program, which allowed the multidisciplinary team assembled during the QI project to be sustained. This new program is seeking means of solidifying the gains from the existing QI process and investigating new ways of achieving further improvement for early PM&R, including designing new medical devices to assist with ambulating mechanically ventilated patients and implementing or evaluating other evidence-based rehabilitation interventions, such as cycle ergometry and neuromuscular electrical stimulation therapy. 33,45-47 Moreover, as of July 2009, the approach to sedation that was encouraged during the QI project has been formalized as a new treatment protocol, and standardized delirium evaluation has been implemented as a routine nursing assessment throughout several ICUs at 2 of our hospitals. Limitations This QI project has limitations. First, given its design as a QI project with a before-after comparison, patients were not randomized to sedation or PM&R interventions, nor were the outcomes evaluated in a blinded manner. Hence, the results may be subject to measurement bias and temporal changes. However, the purpose of this project was not to test the efficacy of these interventions, because there are previously published studies demonstrating the safety, feasibility, and benefits of these activities, but to undertake a structured QI process to determine if routine clinical practice could be substantially and rapidly improved. Such a change may not be easy given that it requires a significant transformation in culture for the entire multidisciplinary ICU team, which can be extremely difficult to achieve in a relatively short time frame. 40 Second, given the small size and duration of this QI project and its focus in a single MICU in an academic teaching hospital, the results may not be generalizable to other types of ICUs or hospitals. This limitation is central to the vast majority of QI work that is conducted at single institutions. However, in accordance with recent guidelines on the reporting of QI projects, 29 we have described both the QI process and interventions in detail in order to assist with understanding the specific context of our MICU and customizing the process and intervention for replication at other institutions. Finally, as a multifaceted QI project, we cannot determine which specific aspects (eg, changed sedation vs early PM&R) may be associated with the reduction in LOS, especially because prior literature for both interventions has demonstrated such beneficial effects. 13,14,19,22,40,41 CONCLUSIONS Using a structured and multifaceted QI process, we quickly and markedly reduced the use of deep sedation and increased early PM&R activities for mechanically ventilated patients. Through these activities, substantial improvements were observed in ICU delirium and patients functional mobility, with a decrease in MICU and hospital LOS. References 1. De Jonghe B, Sharshar T, Lefaucheur JP, et al. Paresis acquired in the intensive care unit: a prospective multicenter study. JAMA 2002;288:2859-67. 2. Fletcher SN, Kennedy DD, Ghosh IR, et al. Persistent neuromuscular and neurophysiologic abnormalities in long-term survivors of prolonged critical illness. Crit Care Med 2003;31:1012-16. 3. Herridge MS, Cheung AM, Tansey CM, et al. One-year outcomes in survivors of the acute respiratory distress syndrome. N Engl J Med 2003;348:683-93. 4. Weinert CR, Calvin AD. Epidemiology of sedation and sedation adequacy for mechanically ventilated patients in a medical and surgical intensive care unit. Crit Care Med 2007;35:393-401. 5. Goldhill DR, Badacsonyi A, Goldhill AA, Waldmann C. A prospective observational study of ICU patient position and frequency of turning. Anaesthesia 2008;63:509-15. 6. Needham DM. Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA 2008;300:1685-90. 7. Berg HE, Larsson L, Tesch PA. Lower limb skeletal muscle function after 6 wk of bed rest. J Appl Physiol 1997;82:182-8. 8. Kortebein P, Ferrando A, Lombeida J, Wolfe R, Evans WJ. Effect of 10 days of bed rest on skeletal muscle in healthy older adults. JAMA 2007;297:1772-4. 9. Dowdy DW, Eid MP, Dennison CR, et al. Quality of life after acute respiratory distress syndrome: a meta-analysis. Intensive Care Med 2006;32:1115-24. 10. Hamburg NM, McMackin CJ, Huang AL, et al. Physical inactivity rapidly induces insulin resistance and microvascular dysfunction in healthy volunteers. Arterioscler Thromb Vasc Biol 2007;27: 2650-6. 11. Clavet H, Hebert PC, Fergusson D, Doucette S, Trudel G. Joint contracture following prolonged stay in the intensive care unit. CMAJ 2008;178:691-7. 12. Bailey P, Thomsen GE, Spuhler VJ, et al. Early activity is feasible and safe in respiratory failure patients. Crit Care Med 2007;35: 139-45. 13. Morris PE, Goad A, Thompson C, et al. Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med 2008;36:2238-43. 14. Schweickert WD, Pohlman MC, Pohlman AS, et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet 2009;373:1874-82. 15. Gosselink R, Bott J, Johnson M, et al. Physiotherapy for adult patients with critical illness: recommendations of the European Respiratory Society and European Society of Intensive Care Med-

542 EARLY PHYSICAL MEDICINE AND REHABILITATION FOR ACUTE RESPIRATORY FAILURE, Needham icine Task Force on Physiotherapy for Critically Ill Patients. Intensive Care Med 2008;34:1188-99. 16. Needham DM, Dennison CR, Dowdy DW, et al. Study protocol: The Improving Care of Acute Lung Injury Patients (ICAP) study. Crit Care 2005;10:R9 http://ccforum.com/content/10/1/r9. March 6, 2010. 17. Needham DM, Wang W, Desai SV, et al. Intensive care unit exposures for long-term outcomes research: development and description of exposures for 150 patients with acute lung injury. J Crit Care 2007;22:275-84. 18. Zanni JM, Korupolu R, Fan E, et al. Rehabilitation therapy and outcomes in acute respiratory failure: An observational pilot project. J Crit Care. In press. 19. Kress JP, Pohlman AS, O Connor MF, Hall JB. Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342:1471-7. 20. Pronovost PJ, Berenholtz SM, Needham DM. Translating evidence into practice: a model for large scale knowledge translation. BMJ 2008;337:a1714. 21. Treggiari MM, Romand JA, Yanez ND, et al. Randomized trial of light versus deep sedation on mental health after critical illness. Crit Care Med 2009;37:2527-34. 22. Brook AD, Ahrens TS, Schaiff R, et al. Effect of a nursingimplemented sedation protocol on the duration of mechanical ventilation. Crit Care Med 1999;27:2609-15. 23. Ely EW, Truman B, Shintani A, et al. Monitoring sedation status over time in ICU patients: reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 2003;289:2983-91. 24. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373-83. 25. American College of Critical Care Medicine and American Society of Health-System Pharmacists. Clinical practice guidelines for the sustained use of sedatives and analgesics in the critically ill adult. Am J Health Syst Pharm 2002;59:150-78. 26. Wilson WC, Smedira NG, Fink C, McDowell JA, Luce JM. Ordering and administration of sedatives and analgesics during the withholding and withdrawal of life support from critically ill patients. JAMA 1992;267:949-53. 27. Ely EW, Inouye SK, Bernard GR, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA 2001;286:2703-10. 28. Rabe-Hesketh S, Skrondal A. Multilevel and longitudinal modeling using Stata. College Station: Stata Pr; 2005. 29. Davidoff F, Batalden P, Stevens D, Ogrinc G, Mooney S. Publication guidelines for quality improvement in health care: evolution of the SQUIRE project. Qual Saf Health Care 2008;17(Suppl 1):i3-9. 30. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006;355:2725-32. 31. Ferrer R, Artigas A, Levy MM, et al. Improvement in process of care and outcome after a multicenter severe sepsis educational program in Spain. JAMA 2008;299:2294-303. 32. Berenholtz SM, Pronovost PJ, Lipsett PA, et al. Eliminating catheter-related bloodstream infections in the intensive care unit. Crit Care Med 2004;32:2014-20. 33. Korupolu R, Gifford J, Needham DM. Early mobilization of critically ill patients: reducing neuromuscular complications after intensive care. Contemp Crit Care 2009;6(9):1-12. 34. Herridge MS. Mobile, awake and critically ill. CMAJ 2008;178: 725-6. 35. Morris PE. Moving our critically ill patients: mobility barriers and benefits. Crit Care Clin 2007;23:1-20. 36. Bergel RR. Disabling effects of inactivity and importance of physical conditioning. A historical perspective. Rheum Dis Clin North Am 1990;16:791-801. 37. Dock W. The evil sequelae of complete bed rest. JAMA 1944; 125:1085. 38. Petty TL. Suspended life or extending death? Chest 1998;114: 360-1. 39. Burns JR, Jones FL. Early ambulation of patients requiring ventilatory assistance [letter]. Chest 1975;68:608. 40. Hopkins RO, Spuhler VJ, Thomsen GE. Transforming ICU culture to facilitate early mobility. Crit Care Clin 2007;23:81-96. 41. De Jonghe B, Bastuji-Garin S, Fangio P, et al. Sedation algorithm in critically ill patients without acute brain injury. Crit Care Med 2005;33:120-7. 42. Riker RR, Fraser GL. Altering intensive care sedation paradigms to improve patient outcomes. Crit Care Clin 2009;25:527-38, viii-ix. 43. Riker RR, Shehabi Y, Bokesch PM, et al. Dexmedetomidine vs midazolam for sedation of critically ill patients: a randomized trial. JAMA 2009;301:489-99. 44. Kress JP, Gehlbach B, Lacy M, Pliskin N, Pohlman AS, Hall JB. The long-term psychological effects of daily sedative interruption on critically ill patients. Am J Respir Crit Care Med 2003;168: 1457-61. 45. Burtin C, Clerckx B, Robbeets C, et al. Early exercise in critically ill patients enhances short-term functional recovery. Crit Care Med 2009;37:2499-505. 46. Truong AD, Fan E, Brower RG, Needham DM. Bench-to-bedside review: mobilizing patients in the intensive care unit from pathophysiology to clinical trials. Crit Care 2009;13:216. 47. Needham DM, Truong AD, Fan E. Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med 2009; 37:S436-41. Supplier a. Stata Corporation, 4905 Lakeway Dr, College Station, TX 77845.