|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Division of Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, WA.
Correspondence to: Gordon D. Rubenfeld, MD, MSc, Division of Pulmonary and Critical Care, Harborview Medical Center, Box 359762, 325 Ninth Ave, Seattle, WA 98104; e-mail: nodrog{at}u.washington.edu
| Abstract |
|---|
|
|
|---|
Design: Population-based cross-sectional study.
Setting: Seventeen adult hospitals in King County, WA.
Participants/interventions: We compared the blank RT ICU flow sheet for each institution to the 1992 American Association for Respiratory Care (AARC) clinical practice guidelines (CPGs) for patient-ventilator system checks. We interviewed RT managers at each hospital about their practices. Finally, we reviewed selected charts of patients with acute lung injury (ALI) or ARDS from each hospital to evaluate the documentation.
Measurements/results: We found substantial variability in RT documentation practices and in their extent of compliance with the AARC CPGs. Only 15 of 52 items recommended by the AARC CPGs were included on blank RT flow sheets of every hospital in our study, and only 26 of 52 items were found on charts of ALI/ARDS patients at most hospitals (ie,
10 of 17 hospitals). Only 10 of 17 RT department managers reported using the AARC CPGs as a basis for their documentation policies. Items necessary for the implementation of lung-protective ventilation for ALI/ARDS patients were recorded inconsistently and were not included in the AARC CPGs. Plateau pressure was found on all reviewed charts of ALI/ARDS patients at only 10 of 17 hospitals.
Conclusions: Considerable variability exists in RT documentation practices. We suggest that new guidelines be developed for documenting the care of patients receiving mechanical ventilation, in light of recent data on ventilator weaning and the management of ALI/ARDS, and that their effect on practice and outcomes be evaluated.
Key Words: ARDS clinical practice guidelines documentation mechanical ventilation respiratory therapy
| Introduction |
|---|
|
|
|---|
Some studies have demonstrated the benefits of respiratory therapy (RT)-directed protocols in weaning patients from mechanical ventilation and of protocols for lung-protective ventilation for acute lung injury (ALI) and ARDS.13 14 15 16 17 The implementation and monitoring of protocols require appropriate documentation and communication of ventilator settings by respiratory therapists to other members of the health-care team. These practices have not been studied.
In 1992, the American Association for Respiratory Care (AARC) published CPGs for patient-ventilator system checks by respiratory therapists.18 To our knowledge, no other CPGs for patient-ventilator checks have been published or promulgated by a national professional organization. In the AARC CPGs, patient-ventilator checks were defined as "a documented evaluation of a mechanical ventilator and of the patients response to mechanical ventilatory support." These CPGs note that "observations should include but are not limited to" the 52 items listed in Table 1 . They recommend that patient-ventilator system checks be performed on a scheduled basis but leave exact frequency to the discretion of each institution.
|
We conducted this study as a first step toward understanding and improving RT, physician, and nurse interactions in the ICU. We examined the variation in RT documentation practices in ICUs, measured the extent of compliance with current recommendations for the documentation of mechanical ventilation, and assessed whether respiratory therapists routinely document the variables necessary to perform lung-protective ventilation for ALI/ARDS. Because responses to surveys may differ from actual practice, we used the following three sources for this study: questionnaire; review of documentation forms; and chart review of ALI/ARDS patients receiving mechanical ventilation.
| Methods and Materials |
|---|
|
|
|---|
Hospitals
Twenty-one hospitals involved in the King County Lung Injury Project (KCLIP) between April 1999 and July 2000 were the initial sample for this study. KCLIP is a population-based cohort study that is designed to measure ALI/ARDS incidence. These 21 hospitals account for 98.1% of all King County residents who have been admitted for mechanical ventilation. Two pediatric hospitals were excluded because of their unique patient populations, and two adult hospitals were dropped because they cared for too few patients with ALI/ARDS, according to our chart review criteria. Data collection was conducted between September 2001 and March 2002.
Data Sources
Flow Sheets:
Blank RT flow sheets for mechanically ventilated patients were compared to the 52 items recommended by the AARC CPGs (Table 1) . Printouts of standard flow sheets were used for the four hospitals that employed a computerized data collection system. (Note that none of the hospitals utilize personal digital assistants for data entry by respiratory therapists.) We also assessed the documentation of five additional items (Table 1)
that provide information necessary for implementation of the National Institutes of Health ARDS Network low-tidal volume (VT) protocol16
(ie, plateau pressure [Pplat] and weight) or that may be otherwise useful for practitioners (ie, total minute ventilation [
E]). Note that there was no information available on the technique used to measure Pplat at each hospital. Items were coded as "present" or "absent."
Chart Review:
KCLIP identified all patients with ALI/ARDS over a 1-year period at each study hospital. ALI/ARDS was defined as follows, using the American-European Consensus Conference criteria: PaO2/fraction of inspired oxygen (FIO2) ratio of < 300 mm Hg, bilateral radiographic opacities, and no evidence of left atrial hypertension.21
We further restricted the KCLIP sample to patients who survived at least 72 h and required FIO2 of
0.60 at some point during the first 3 days after the onset of ALI/ARDS. From this, we selected five ALI/ARDS patients from each hospital (less than five patients if fewer patients fit our selection criteria) who had the lowest PaO2/FIO2 ratios at the time of meeting the criteria for ALI/ARDS. We selected these charts to identify a subset of patients with ALI/ARDS who were sufficiently ill and had required mechanical ventilation long enough to have ALI/ARDS-specific ventilator parameters recorded if the respiratory therapists at that hospital record them in any patient.
For each patient, we reviewed all RT documentation (ie, ventilator flow sheets and other interdisciplinary flow sheets, progress notes, and medication sheets) for the first 72 h of mechanical ventilation after meeting the study criteria for ALI/ARDS. The total number of checks of each of the 57 items under study during this time period was recorded.
For each hospital, the number of checks and the related number of patient-hours for each item then were summed. The results are reported as whether an item was ever checked on any patient, whether an item was checked on every patient chart that was reviewed, and the mean and range of frequency of checks. (Note that frequency of documentation was only calculated for the items that were appropriate for the ventilation mode. For instance, if a patients ventilator was operating on pressure support (PS) mode, that period of time was excluded from the measurement of frequency of static pressure or auto-positive end-expiratory pressure [PEEP] checks.)
RT Manager Survey: We conducted a telephone survey with the RT manager who was responsible for the ICUs at each hospital (Table 2 ). At two hospitals, the RT manager deferred the interview to a senior ICU therapist. The survey consisted of 3 general questions (ie, number of ICU beds, number of therapists assigned to the ICU, and the available ventilator types) and 11 yes/no questions about patient-ventilator system checks.
|
All data collection was performed by a single person. Random repeat extraction was performed on 10% of the charts with good agreement.
| Results |
|---|
|
|
|---|
|
|
E; arterial blood gas (ABG) values or oxygen saturation; weaning parameters; endotracheal tube (ETT) size and position; ETT cuff pressure; inspiratory gas temperature; time of last patient circuit change; and description of patient secretions. Finally, a total of 17 items were infrequently (ie, in < 10 hospitals) present on blank flow sheets.
|
On chart review, respiratory therapists at all 17 hospitals consistently documented identification information such as patient name, patient hospital number, date and time of current check, and respiratory therapist signature or initials. Ventilator settings that were consistently recorded in all charts at 16 of 17 hospitals included the following: FIO2 (set FIO2, 16 hospitals; only analyzed FIO2, 1 hospital); mode; set rate; alarm settings; PEEP; total
E; peak airway pressure; and inspiratory flow rate. The clinical observations that were always recorded were the patients oxygenation status (ie, ABG levels or saturation) and patients spontaneous respiratory rate (or total number of respirations above the set ventilator rate).
None of the chart sections reviewed documented patient diagnosis, presence or absence of an instance of equipment failure, check of the ventilator being on, check of the patient circuit being secure, check of the disconnect alarm, or condition of ancillary equipment. Sigh variable was noted on charts at two hospitals only. Documentation of the operational verification procedure being performed was found only on one chart in one hospital. Finally, the presence or absence of untoward effects of disconnection from the ventilator during bedside procedures was noted at five hospitals, but not on every chart. Comment on patient-ventilator synchrony was found at two hospitals but not on every chart.
Documentation of the rest of the items was variable. Most, but not all, hospitals recorded set peak inspiratory pressure limit, set minimum mandatory
E (ie, the low
E alarm), ETT size, ETT position, ETT cuff pressure, PS, set VT, delivered VT, spontaneous VT, inspired gas temperature, breath sounds, and description of patient secretions in all patients. The remaining 20 items (Table 1)
were only infrequently (ie, < 10 of 17 hospitals) included on every chart.
The documentation practices of several of the most important items are depicted in Table 5 . These include ventilator settings relevant to lung-protective ventilation strategies (ie, FIO2 and airway pressures) and those related to patient safety (ie, inspired gas temperature, ETT cuff pressure, and others). Recording practices ranged from items never recorded in every patient (ie, VT in milliliters per kilogram), those infrequently recorded in every patient (ie, auto-PEEP [3 of 17 hospitals always recorded]), to those usually being recorded (ie, PEEP [16 of 17 hospitals]). When recorded, the frequency of documentation of these items also varied greatly. For example, set FIO2 was recorded a mean of 10.6 times per 24 h (range, 6.4 to 16.5 times), auto-PEEP was recorded 2.9 times (range, 0.1 to 5.6 times), and time of last patient circuit change was recorded 0.5 times (range, 0.1 to 2.1 times).
|
|
| Discussion |
|---|
|
|
|---|
10 of 17 hospitals). The extent of documentation of the remaining items varied widely across hospitals, and this is particularly noticeable in the differences in the frequency of checks of various items. With respect to the items that were assessed, there was inconsistency of documentation practice within each hospital. Inclusion of an item on the blank RT flow sheet did not assure the charting of this item on every patient. In addition, there were items (eg, set VT, auto-PEEP, and ETT cuff pressure) that were recorded on some but not all of the charts of ALI/ARDS patients reviewed at each hospital. There was also discrepancy between the practice reported by the RT managers and the actual charting. Conversely, although some items were excluded from the blank RT flow sheet, they were written in on some of the patients charts. This was true not only for clinical observations of the patient and comments on adverse events, but also for specific ventilator settings and measures such as delivered VT and the ratio of inspiratory to expiratory time.
When we focused on key items that are essential for the implementation of lung-protective ventilation for ALI/ARDS patients, we found that these were absent from many charts. Four of the 17 hospitals did not include Pplat on their blank RT flow sheets. Five of the 17 hospitals did not record Pplat on any of the reviewed ALI/ARDS patients charts. Two additional hospitals did not record Pplat on every ALI/ARDS patient chart reviewed. Sixteen of the 17 hospitals had no notation for VT in milliliters per kilogram on any reviewed ALI/ARDS patients chart.
The reasons for the marked variability in practice within and across hospitals, and for the failure to record key measures are not known. These could be related to the hospitals experience with mechanical ventilation. Despite the differences in the numbers of ICU beds, the number of respiratory therapists assigned to the ICU, or the number of mechanical ventilation cases seen per year at each hospital, grouped analysis of our results by type of hospital (ie, university-affiliated, community academic, and community nonacademic) [data not shown] demonstrated similar results. However, our survey included insufficient numbers of different types of hospitals to demonstrate a difference in practice based on university affiliation or size.
The lack of updated and easily applicable documentation guidelines may be a reason for the variability in practice. The AARC CPGs for performance and documentation of the patient-ventilator system check were published in 1992 as part of a larger AARC initiative to develop CPGs. Review of literature, practice surveys, and expert panel opinion were used to draft the CPGs. Revisions were based on input from AARC working groups (led by national experts in respiratory care) and up to 700 independent reviewers.22 Although the CPGs were an innovative and important development for RT, there was no further study of the impact of these guidelines on practice or outcomes, and the guidelines have not been revised or updated. This may be why, as our study suggests, the CPGs are not in widespread use today. Seven of 17 RT department managers reported that the AARC CPGs were not used in the design of their own hospitals RT documentation policies. This may imply that these guidelines are not an accurate reflection of current RT practices and beliefs about documentation. It also may explain our findings of substantial discordance between the AARC CPGs and the RT ICU documentation practices.
Only 15 of 52 items recommended by the AARC CPGs were included on the blank RT flow sheets of every hospital in our study, and only 26 of 52 items were found on patient charts in most hospitals (ie,
10 of 17 hospitals). Factors such as limited staffing and time, and the questionable significance of some of the items for RT assessment may be some of the reasons for failure to comply strictly with the CPGs. It would be time-consuming and redundant to document all routine safety tests at every patient-ventilator system check, as the CPGs recommend. Similarly, noting and charting a patients skin color or status of chest tubes may not be as relevant for the respiratory therapist as for other members of the health-care team.
Another problem may be that the extensive list of items recommended by the CPGs fails to include some measures that are considered essential in current practice. These guidelines were published before RT-driven weaning protocols and lung-protective ventilation strategies were shown to improve outcomes. The 1992 AARC CPGs do not recommend assessing and documenting total
E, which may be useful in the assessment of work of breathing and readiness for weaning.13
Even more important is Pplat. Not only is this an accurate reflection of alveolar pressures and risk of barotrauma,23
it is essential for the implementation of lung-protective ventilation for ALI/ARDS patients.15
16
The CPGs do not recommend assessing and documenting Pplat.
This study has several potential limitations. First, we only examined charting practices. It is possible that some information was collected and conveyed outside of the available chart documentation. Although documentation may not reflect all the care being delivered, it was the only available means to assess guideline adherence. No conclusions can be drawn about the relation of adherence to these recommended documentation practices and the quality of care provided at these hospitals. However, documentation is a key means of communication between respiratory therapists and other members of the medical team, and comprehensive charting of relevant information may be essential for efficient and appropriate ventilator management. Second, only a small number of patient charts were reviewed from each hospital. We attempted to select the charts of the most severely ill ALI/ARDS patients in order to bias toward the most accurate and complete documentation. Therefore, we believe that the results reflect the highest level of RT charting found at each hospital. Third, because we examined the practices in a limited number of hospitals in a specific geographic area, our results may not be broadly generalizable. We are unable to draw conclusions regarding practice at specific types of hospitals or to make convincing comparisons among these groups. We also did not have the information needed to assess differences in documentation practices between medical and surgical ICUs. Fourth, the measurement of some items may be easier with specific ventilators, and this may bias practice. Although we noted the types of ventilators that were generally available at the surveyed hospitals, we did not correlate documentation practice with ventilator type. Finally, this was a cross-sectional measure of current practice, and it did not address the greater question of the impact of RT documentation practices and CPGs on clinical outcomes.
We believe that there is a need for a focus on RT documentation practices in the ICU and that this study is a first step in this process. Our findings suggest that greater efforts should be made to standardize RT documentation processes, to revise and update CPGs and RT flow sheets, and to investigate the impact of RT documentation on clinical outcomes in the ICU.
| Conclusions |
|---|
|
|
|---|
| Footnotes |
|---|
E = minute ventilation; VT = tidal volume This research was supported by grants SCOR HL30542 and RO1 HL67939 from the National Institutes of Health, Bethesda, MD.
Received for publication January 22, 2003. Accepted for publication April 21, 2003.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |