The administration of respiratory treatments is a

The Effect of Respiratory

Therapist-Initiated Treatment Protocols on Patient

Outcomes and Resource Utilization*

Marin H. Kollef, MD, FCCP; Steven D. Shapiro, MD, FCCP;

Darnetta Clinkscale, MA; Lisa Cracchiolo, RRT; Donna Clayton, BS;

Russ Wilner, RRT; and Linda Hossin, RRT

Context:

Physicians frequently prescribe respiratory treatments to hospitalized patients, but the

influence of such treatments on clinical outcomes is difficult to assess.

Objective:

To compare the clinical outcomes of patients receiving respiratory treatments

managed by respiratory care practitioner (RCP)– directed treatment protocols or

physician-directed orders.

Design:

A single center, quasi-randomized, clinical study.

Setting:

Three internal medicine firms from an urban teaching hospital.

Patients:

Six hundred ninety-four consecutive hospitalized non-ICU patients ordered to receive

respiratory treatments.

Main outcome measures:

Discordant respiratory care orders, respiratory care charges, hospital

length of stay, and patient-specific complications. Discordant orders were defined as written

orders for respiratory treatments that were not clinically indicated as well as orders omitting

treatments that were clinically indicated according to protocol-based treatment algorithms.

Results:

Firm A patients (n

ⴝ

239) received RCP-directed treatments and had a statistically lower

rate of discordant respiratory care orders (24.3%) as compared with patients receiving

physician-directed treatments in firms B (n

ⴝ

205; 58.5%) and C (n

ⴝ

250; 56.8%; p

<

0.001). No

statistically significant differences in patient complications were observed. The average number

of respiratory treatments and respiratory care charges were statistically less for firm A patients

(10.7

ⴞ

13.7 treatments; $868

ⴞ

1,519) as compared with patients in firms B (12.4

ⴞ

12.7

treatments, $1,124

ⴞ

1,339) and C (12.3

ⴞ

13.4 treatments, $1,054

ⴞ

1,346; p

ⴝ

0.009

[treat-ments] and p

<

0.001 [respiratory care charges]).

Conclusions:

Respiratory care managed by RCP-directed treatment protocols for non-ICU

patients is safe and showed greater agreement with institutional treatment algorithms as compared

with physician-directed respiratory care. Additionally, the overall utilization of respiratory treatments

was significantly less among patients receiving RCP-directed respiratory care.

(CHEST 2000; 117:467–475)

Abbreviations: APACHE⫽acute physiology and chronic health evaluation; BJH⫽Barnes-Jewish Hospital; CI⫽confidence interval; RCP⫽respiratory care practitioner

T

he administration of respiratory treatments is a

necessary component of the care of hospitalized

patients, especially those individuals with

thoracoab-dominal disease processes. However, a number of

clinical studies suggest that respiratory treatments are

often needlessly prescribed to patients who do not

derive benefit from their administration.

_{To date,}

few investigations have attempted to define the optimal

method for the utilization of respiratory treatments or

to determine the impact of various respiratory care

utilization strategies on patient outcomes.

_This

represents an important area of investigation because

of the potential complications associated with

inappro-priate use of respiratory treatments and the need to

conserve valuable medical resources for treatments

that are more likely to provide patient benefits.

To better determine the optimal use of respiratory

*From the Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Washington University School of Medicine (Drs. Kollef and Shapiro), and Department of Respi-ratory Care Services, Barnes-Jewish Hospital (Mss. Clinkscale, Cracchiolo, Clayton, and Hossin, and Mr. Wilner), St. Louis, MO. This investigation was supported by a grant provided by the American Association of Respiratory Care.

Manuscript received March 5, 1999; revision accepted June 28, 1999.

Correspondence to: Marin H. Kollef, MD, FCCP, Pulmonary and Critical Care Medicine, Washington University School of Medi-cine, Campus Box 8052, 660 South Euclid, St. Louis, MO 63110; e-mail: mkollef@pulmonary.wustl.edu

treatments for hospitalized patients not requiring

ICU admission (

ie

, non-ICU patients), we

per-formed a clinical trial that had two main objectives.

The first goal of our study was to determine the

occurrence of administered respiratory treatments

that were discordant with the respiratory care

guide-lines of the hospital. Our second goal was to test the

hypothesis that the use of respiratory care

practi-tioner (RCP)-directed treatment protocols would

decrease the occurrence of discordant respiratory

treatments.

Materials and Methods

Study Location and Population

The study was conducted at a university-affiliated urban teaching hospital: Barnes-Jewish Hospital (BJH; 1,400 beds). During a 9-month period (February 1998 to October 1998), all patients ordered to receive respiratory treatments from the three internal medicine firms of the hospital (firms A, B, C) were potentially eligible for this investigation. The internal medicine firms are closed groups of internal medicine house officers, teaching physicians, and private attending physicians caring for patients on the medicine floors of BJH. Physicians are assigned to the firms so as to achieve a similar distribution of teaching and private attending physicians as well as a similar distribution of general internists and subspecialty-trained physicians, including pulmonary subspecialists. The patient populations served by the three firms are also similar in terms of their physicians’ practices (eg, general internal medicine vs subspecialty practices). There is no crossover of physicians between the internal medicine firms, which also have their own administrative structures (eg, chief residents, chiefs of service). Patients admitted to the internal medicine firms received their care on one of four hospital floors according to bed availability. Each of the medical floors involved in this study had a mixture of patients from the different medical firms. Patients were entered into the investigation if they were

⬎16 years old and excluded if they had a diagnosis of cystic fibrosis. The study was approved by the Washington University School of Medicine Human Studies Committee.

Study Design

Patients were assigned, at the time of hospital admission, to have their respiratory care managed by RCP-directed treatment protocols or by physician-directed orders. All patients assigned to firm A received RCP-directed treatments, whereas patients assigned to firms B and C received physician-directed respiratory care. Patient assignment to a medicine firm was based on their attending or clinic physician’s firm assignment. All patients without an attending or clinic physician were randomly assigned by the hospital admissions office to one of the medicine firms. The primary outcome measure was the occurrence of discordant respiratory care orders. Discordant respiratory care orders were defined as written orders that omitted clinically indicated treat-ments (eg, absence of an order for bronchopulmonary hygiene in a patient with lobar atelectasis), as well as orders for treatments that were not clinically indicated according to protocol-based treatment algorithms (eg, an order for bronchopulmonary hy-giene in a patient without an appropriate indication; Fig 1). Secondary outcome measures included the total number of respiratory care treatments per patient, charges for respiratory

care treatments, hospital length of stay, patient-specific compli-cations, and hospital mortality. All data were collected on a standardized data collection form by registered respiratory ther-apists who were blinded to patients’ firm assignments as well as to whether patients’ respiratory care orders were protocol guided. Respiratory Care Treatment Protocols

A series of respiratory care treatment protocols have been developed at BJH using a branched-chain logic format similar to that used at the Cleveland Clinic.12,13,16_{These protocols follow} the recommendations of the American Association for Respira-tory Care Clinical Practice Guidelines.17,18 _{The protocols were} developed by a working group of physicians and RCPs at BJH. Before their acceptance, each protocol underwent a trial period of utilization to determine its safety and effectiveness. The protocols were subsequently published as a hospital resource guide and made available to all physicians with admitting privi-leges to BJH, hospital nursing units, and hospital administrative staff. Additionally, each house officer and RCP was given a pocket-sized version of the resource guide before beginning the investigation. All RCPs underwent educational in-service training focused on the performance of patient assessments and the utilization of the treatment protocols. An example of a RCP treatment protocol is provided in Figure 1.

Two RCPs (L.C. and L.H.) prospectively evaluated all respi-ratory care orders to determine the presence of discordant orders. These two RCPs were blinded to patients’ firm assign-ments as well as to whether patients’ respiratory care orders were protocol guided, had no other clinical or administrative duties during their participation in the study, and contributed to the initial development and updating of the respiratory care treat-ment protocols used at BJH. Respiratory care orders were obtained daily by the two reviewers using an automated medical order communication system without having to directly review the medical record order sheets. This was done to minimize observer bias in regards to who wrote the original respiratory care orders (physician vs respiratory therapist). All respiratory care orders, discordant order assessments, and patient outcomes were subsequently reviewed by the medical directors of the respiratory care department (M.H.K. and S.D.S.). This was done to ensure that the discordant classification of respiratory care orders ad-hered to the explicit definitions described above.

RCP-Directed Respiratory Care

Each firm A patient ordered to receive RCP-directed treat-ments underwent formal assessment by a registered respiratory therapist who had⬎5 years of experience as a respiratory care supervisor. The respiratory therapist performing the patient assessment determined the patient’s respiratory care needs based on the institutional treatment protocols. A respiratory care plan was developed and orders were written. The patient’s treating physician requesting respiratory care was informed of the patient assessment. Treating physicians who disagreed with the assess-ment and orders of the RCP performing the evaluation were allowed to change the orders. However, all disagreements were subsequently reviewed by the medical directors of the respiratory care department for final discussion with the treating physicians. Physician-Directed Respiratory Care

All aspects of respiratory care were ordered by the treating physicians for patients assigned to firms B and C. Respiratory therapists could not make changes in the treatment orders without a physician’s order. However, respiratory therapists could

communicate their opinions and observations about patients’ needs for respiratory care to the treating physicians.

Definitions

All definitions were selected prospectively as part of the original study design. We calculated acute physiology and chronic health evaluation (APACHE) II scores on the basis of clinical data available from the first 24-h period of hospitalization.19 Lifestyle scores were previously defined as follows20_{: 0 indicates} employed; 1, independent, fully ambulatory; 2, restricted activi-ties, able to live on own and get out of home to do basic necessities, severe limitation in exercise ability; 3, housebound, cannot get out of house unassisted, cannot live alone or perform heavy chores; and 4, bed- or chair-bound. The diagnostic criteria for nosocomial pneumonia were modified from those established by the American College of Chest Physicians.21 _Nosocomial

pneumonia was defined as the development of a new or progres-sive radiographic infiltrate in conjunction with two of the follow-ing: fever (temperature⬎38.3°C), leukocytosis (leukocyte count

⬎10⫻109_{/L), and purulent tracheal aspirate. Lobar atelectasis} was defined as displacement of a fissure along with opacification of a lobe demonstrated radiographically. Immunosuppression was defined as patients receiving corticosteroids, having a positive serum HIV antibody, having received chemotherapy in the past 45 days, or having neutropenia (absolute neutrophil count

⬍0.5⫻109_{/L) resulting from the administration of} chemother-apy, or as recipients of an organ transplant (kidney, liver, heart, lung, bone marrow) requiring immunosuppressive agents. Statistical Analysis

We estimated the sample size needed to provide 90% power to detect a difference in the rate of discordant orders of 20%. We Figure1. Treatment guideline for bronchial hygiene. IPPB⫽intermittent positive-pressure

used an␣error of 0.05 (two-tailed) and assumed a baseline rate of discordant orders of 50% on the basis of preliminary surveys conducted before performing the investigation. According to these assumptions, 269 patients were needed in each study group to provide the desired power. All comparisons were unpaired, and all tests of significance were two-tailed. Continuous variables were compared using the Student’sttest for normally distributed variables and the Wilcoxon rank sum test for non-normally distributed variables. The␹2_{or Fisher’s Exact Test were used to} compare categorical variables. The primary data analysis was an intention-to-treat analysis, comparing the rate of discordant orders among patients assigned to receive RCP-directed treat-ments in firm A and the patients receiving physician-directed respiratory care in firms B and C.

Multiple logistic regression analysis, using a commercial statis-tical package, was used to identify predictor variables that were significantly related to the likelihood of having a discordant order for respiratory care (eg, presence of a discordant order as the dependent outcome variable).22 _{Baseline covariants were} in-cluded in models that were judgeda priorito be clinically sound. This was prospectively determined to be necessary to avoid producing spuriously significant results with multiple compari-sons.23_{Potential predictor variables for model entry were} iden-tified using univariate analysis, where a p value of 0.15 was used to determine entry into the logistic regression model. A stepwise

approach was used to enter new terms into the logistic regression model, and 0.05 was set as the limit for the acceptance or removal of these terms. Results of the logistic regression analysis are reported as adjusted odds ratios with 95% confidence intervals (CIs).

Results

Patients

Six hundred ninety-four consecutive patients

or-dered to receive respiratory treatments from firms A,

B, and C were enrolled into the study. At the time of

randomization, no statistically significant differences

were found among the three firms for age, sex,

ethnicity, lifestyle scores, APACHE II scores,

smok-ing status, presence of a do-not-resuscitate order,

tracheostomy at the time of hospital admission, and

the presence of COPD, congestive heart failure,

underlying malignancy, seropositivity for HIV, or an

immunocompromised state (Table 1). Patients in

firm B were statistically more likely to have an

Table 1—Patient Characteristics at Time of Hospital Admission*

Characteristics Firm A (n⫽239) Firm B (n⫽205) Firm C (n⫽250) p Value Age, yr 60.8⫾17.5 61.5⫾17.8 59.7⫾17.2 0.366 Male 96 (40.2) 82 (40.0) 108 (43.2) 0.726 Race White 106 (44.4) 85 (41.5) 114 (45.6) 0.574 Black 133 (55.6) 118 (57.5) 134 (53.6) Other 0 (0.0) 2 (1.0) 2 (0.8) Lifestyle score 2.0⫾1.1 1.8⫾1.2 2.0⫾1.1 0.450 APACHE II score 10.5⫾5.1 9.2⫾4.8 10.0⫾4.9 0.086 Smoking status Nonsmoking 68 (28.5) 56 (27.3) 62 (24.8) 0.859 Current smoker 55 (23.0) 53 (25.9) 62 (24.8) Ex-smoker 116 (48.5) 96 (46.8) 126 (50.4) Do-not-resuscitate order 16 (6.7) 8 (3.9) 14 (5.6) 0.433

Congestive heart failure 107 (44.8) 72 (35.1) 102 (40.8) 0.118

COPD 179 (74.9) 167 (81.5) 184 (73.6) 0.117 Underlying malignancy 38 (15.9) 22 (10.7) 30 (12.0) 0.230 HIV-positive 1 (0.4) 2 (1.0) 7 (2.8) 0.096 Immunocompromised 32 (13.4) 24 (11.7) 40 (16.0) 0.406 Tracheostomy 3 (1.3) 2 (1.0) 4 (1.6) 0.916 Admission diagnosis COPD/asthma 58 (24.3) 80 (39.0) 65 (26.0) 0.001

Congestive heart failure 44 (18.4) 26 (12.7) 41 (16.4) 0.254

Pneumonia 30 (12.6) 28 (13.7) 53 (21.2) 0.018 Myocardial ischemia 26 (10.9) 25 (12.2) 24 (9.6) 0.674 Gastrointestinal hemorrhage 12 (5.0) 14 (6.8) 8 (3.2) 0.202 Hepatitis/cirrhosis 1 (0.4) 1 (0.5) 0 (0.0) 0.538 Neurologic 7 (2.9) 3 (1.5) 6 (2.4) 0.567 Sepsis 18 (7.5) 9 (4.4) 13 (5.2) 0.327 Renal failure 6 (2.5) 1 (0.5) 6 (2.4) 0.211 Pulmonary fibrosis 2 (0.8) 0 (0.0) 5 (2.0) 0.128 Cancer 12 (5.0) 2 (1.0) 7 (2.8) 0.042 Pulmonary embolism 4 (1.7) 4 (1.9) 5 (2.0) ⬎0.999 Miscellaneous 19 (8.0) 12 (5.8) 17 (6.8) 0.683

admission diagnosis of COPD or asthma and

statis-tically less likely to have an admission diagnosis of

cancer compared with patients in firms A and C.

Firm C patients were statistically more likely to have

an admission diagnosis of pneumonia compared with

patients in firms A and B.

Respiratory Care Treatments and Discordant

Orders

Patients in firm A received statistically fewer

respiratory treatments as compared with patients in

firms B and C (Table 2; Fig 2). The number of

patients receiving small-volume nebulizers was

sta-tistically less and the number of patients receiving

metered-dose inhalers was statistically greater for

firm A patients as compared with patients in firms B

and C (Table 2; Fig 3). The number of small-volume

nebulizer treatments per treated patient was

signif-icantly less in firm A (9.5

⫾

13.5 treatments) as

compared with patients in firms B (12.3

⫾

12.6

treatments) and C (12.7

⫾

13.2 treatments; p

⬍

0.001).

Conversely, the number of metered-dose inhaler

treat-ments per treated patient was statistically greater for

firm A patients (7.1

⫾

8.4 treatments) as compared

with patents in firms B (5.3

⫾

5.6 treatments) and C

(5.0

⫾

5.1 treatments; p

⫽

0.011).

Three hundred twenty study patients (46.1%) had

at least one discordant order for respiratory care.

Among firm A patients, 58 patients (24.3%) had

discordant orders for respiratory treatments; 262

patients (57.6%) in firms B and C combined had

discordant orders for respiratory treatments (relative

risk, 0.42; 95% CI, 0.33 to 0.53). Among patients

with discordant orders, those in firm A had

statisti-cally fewer discordant orders as compared with

patients in firms B and C (Table 3). When we used

multiple logistic regression analysis to control for all

relevant confounders, the adjusted odds ratio

assess-ing the relationship between the presence of

discor-dant orders and firm assignment (assignment to firm

A as compared with assignment to firms B or C) was

0.50 (95% CI, 0.45 to 0.55). Patients in firm A were

statistically less likely to have a discordant order as

compared with patients in firms B and C in six of the

eight specific categories examined (Table 3).

Physi-cians in firm A changed the orders of respiratory

therapists in 86 instances, which represented 3.4% of

all ordered respiratory treatments for firm A

pa-tients. The medical directors of respiratory therapy

overrode 37 (60.5%) of these physician orders.

Secondary Outcomes

No statistically significant differences in the

occur-rence of nosocomial pneumonia, lobar atelectasis,

pneumothorax, hospital mortality, transfer to an

ICU, or the hospital length of stay were observed

between the three internal medicine firms (Table 4).

Patients in firm A had statistically lower charges for

respiratory care as compared with patients in firms B

and C. Patients with at least one discordant order

had statistically greater charges for respiratory care

as compared with patients having no discordant

orders (Fig 4).

Discussion

Our investigation demonstrated that the use of

RCP-directed protocols significantly reduced the

number of prescribed respiratory treatments as

com-pared with physician-directed respiratory care. Use

of the RCP-directed protocols also significantly

de-creased total charges for respiratory care. These

outcomes were achieved without any increase in

adverse patient outcomes. RCP-directed treatments

were significantly more likely to conform to currently

recommended respiratory care practices at our

insti-tution. The overall rate of discordant respiratory care

orders and the number of discordant orders per

patient were significantly less among patients

receiv-ing RCP-directed treatments as compared with

pa-tients receiving physician-directed respiratory care.

Finally, we found a statistically significant association

between the administration of discordant respiratory

treatments and the charges associated with the

de-livery of respiratory care.

The main importance of our findings is that they

offer an additional strategy for more effectively using

respiratory care in hospitalized patients. Protocols

have previously been used to guide specific aspects

of respiratory care for hospitalized patients including

the weaning of mechanical ventilation,

_infection

control practices for the prevention of

ventilator-associated pneumonia,

_{chest physiotherapy,}

_the

clinical use of arterial blood gases,

_{and inhaled}

bronchodilator administration.

_{Many of those}

protocols were specifically designed for

implemen-tation by RCPs to achieve improvements in patient

outcomes, reduce unnecessary medical care costs,

and to unburden physicians from tasks that could be

performed by RCPs.

_{The use of such protocols}

represents an effective strategy for standardizing

medical practices within medical institutions. The

potential benefits associated with such

standardiza-tion of medical practices include improvements in

patient outcomes, reducing medical care costs,

en-abling the identification of clinical practices

requir-ing change or improvement, improvrequir-ing the quality of

nonblinded clinical research, and establishing a

foundation for the performance of quality

improve-ment efforts.

_{Standardized medical practices and}

protocols achieve such benefits primarily by helping

to create less variable or “chaotic” medical

environ-ments. This occurs by decreasing errors in clinical

management, improving the effectiveness of

avail-able treatments, increasing the accountability of

medical providers, and providing a reference of

measure to assess practices that deviate from the

accepted standard.

The findings of this investigation are consistent

with the results of previous studies demonstrating

the ability of RCPs to prescribe and perform

respi-ratory treatments for hospitalized non-ICU patients.

Stoller and colleagues

_{conducted a randomized}

controlled trial comparing RCP-directed respiratory

care with physician-directed respiratory care. These

investigators found that RCP-directed respiratory

care demonstrated better agreement with the

“stan-dard care plan” of their institution and was associated

with lower costs as compared with

physician-di-rected respiratory care. These outcomes were

achieved without any increase in adverse events,

hospital mortality, or hospital length of stay. Similar

studies have been performed focusing on specific

aspects of respiratory care. Several clinical

investiga-tions have demonstrated the ability of RCPs to

Figure2. Box plots for respiratory treatments administered to patients assigned to the three internal medicine firms. Boxes represent 25th to 75th percentiles with 50th percentile (solid line) and mean (broken line) values shown within the boxes. The 10th and 90th percentiles are shown as capped bars, and closed circles mark the 5th and 95th percentiles.

Figure 3. The percentage of study patients receiving small-volume nebulizer treatments and metered-dose inhaler treat-ments according to their internal medicine firm assigntreat-ments, with upper 95% CIs, are shown.

Table 2—Respiratory Care Services Provided*

Variable Firm A (n⫽239) Firm B (n⫽205) Firm C (n⫽250) p Value Respiratory treatments 10.7⫾13.7 12.4⫾12.7 12.3⫾13.4 0.009 Bronchodilators Small-volume nebulizer 89 (37.2) 122 (59.5) 142 (56.8) ⬍0.001 Metered-dose inhaler 212 (88.7) 157 (76.6) 196 (78.4) 0.001 Oxygen monitoring

Arterial blood gas 13 (5.4) 5 (2.4) 8 (3.2) 0.215

Continuous pulse oximetry 12 (5.0) 6 (2.9) 8 (3.2) 0.435

Intermittent pulse oximetry 1 (0.4) 3 (1.5) 2 (0.8) 0.523

Hyperinflation therapy

Incentive spirometry 5 (2.1) 1 (0.5) 6 (2.4) 0.233

Continuous positive airway pressure 11 (4.6) 16 (7.8) 16 (6.4) 0.372

Intermittent positive-pressure breathing 0 (0.0) 0 (0.0) 1 (0.4) ⬎0.999 Bronchopulmonary hygiene

Percussion and vibration 4 (1.7) 10 (4.9) 7 (2.8) 0.150

Tracheal suctioning 2 (0.8) 0 (0.0) 1 (0.4) 0.645

Supplemental oxygen 117 (49.0) 115 (56.1) 137 (54.8) 0.262

Spirometry 10 (4.2) 5 (2.4) 5 (2.0) 0.319

successfully wean patients from mechanical

ventila-tion as compared with tradiventila-tional physician-directed

weaning.

_{Similarly, Alexander and coworkers}

have shown that simple respiratory care guidelines,

based on clinical practices supported by

peer-re-viewed investigations, could be used to reduce the

performance of unnecessary respiratory treatments

without compromising patient care. Such reductions

were associated with significant cost savings and also

allowed RCPs to concentrate their efforts on

treat-ments that potentially were more likely to result in

beneficial patient outcomes. However, it is

impor-tant to note that most of the institutions involved in

the study of protocols and guidelines for respiratory

care (

eg

, Cleveland Clinic Foundation, Cedars-Sinai

Medical Center) have extensive experience with

protocol development and utilization. Therefore, the

results of clinical investigations examining

protocol-directed therapy from these institutions may, in part,

be related to their medical practice cultures (

ie

,

medical cultures accepting protocol- or

guideline-directed treatments). The importance of such an

environment for the successful implementation of

quality improvement efforts, including

protocol-based therapies, has recently been described.

Our study has several limitations. First, it was

performed within a single institution using house

officers to write most of the treatment orders for

patients admitted to the internal medicine firms.

Therefore, these results may not be generalizable to

patients receiving respiratory care at other centers,

particularly nonteaching hospitals. Second, our

eco-nomic analysis only allowed us to examine

respira-tory care charges and not costs. However, these

charges are directly related to the number and type

of respiratory treatments provided to patients. This is

consistent with other study results showing that the

number of prescribed respiratory treatments is the

most important determinant of respiratory therapy

costs and charges.

_{Reductions in administered}

respiratory treatments may not necessarily reduce

the number of respiratory therapists needed at a

given institution. Therefore, the economic impact of

implemented respiratory therapy protocols should

ideally be assessed in terms of resource utilization as

well as manpower needs. Another important

limita-tion of our study is the possibility of bias on the part

of the RCPs performing the assessments for firm A

patients. It is possible that these RCPs altered their

normal practices to achieve better compliance with

the institutional respiratory care algorithms during

the study period. Lack of compliance with protocols

Variable Firm A (n⫽239) Firm B (n⫽205) Firm C (n⫽250) p Value

Patients with discordant orders 58 (24.3) 120 (58.5) 142 (56.8) ⬍0.001

Discordant orders per patient† 1.2⫾0.4 1.5⫾0.8 1.4⫾0.7 0.022

Discordant order categories

Small-volume nebulizer not indicated 29 (12.1) 68 (33.2) 80 (32.0) ⬍0.001 Inadequate dosing or frequency of inhaled medication 18 (7.5) 36 (17.6) 44 (17.6) ⬍0.001

Inhaled medications not indicated 6 (2.5) 18 (8.8) 23 (9.2) 0.005

Excess dosage of inhaled medication 7 (2.9) 27 (13.2) 27 (10.8) ⬍0.001 Small-volume nebulizer indicated instead of metered-dose inhaler 3 (1.3) 6 (2.9) 2 (0.8) 0.217

Inadequate monitoring of oxygenation 4 (1.7) 13 (6.3) 10 (4.0) 0.034

No indication for IPPB or percussion and vibration 0 (0.0) 6 (2.9) 3 (1.2) 0.015 IPPB or percusion and vibration indicated but not ordered 0 (0.0) 1 (0.5) 3 (1.2) 0.213 *Values given as mean⫾SD or No. (%). See Figure 1 for abbreviation.

†Discordant orders for patients having at least one discordant order identified.

Table 4 —Secondary Clinical Outcomes*

Variable Firm A (n⫽239) Firm B (n⫽205) Firm C (n⫽250) p Value Nosocomial pneumonia 0 (0.0) 2 (1.0) 2 (0.8) 0.466 Lobar atelectasis 9 (3.8) 12 (5.9) 11 (4.4) 0.567 Pneumothorax 0 (0.0) 2 (1.0) 0 (0.0) 0.087 Hospital mortality 5 (2.1) 4 (2.0) 8 (3.2) 0.708 Transfer to ICU 10 (4.2) 10 (4.9) 12 (4.8) 0.927

Hospital length of stay, d 5.9⫾4.1 5.5⫾5.0 6.1⫾3.9 0.489

Respiratory care charges, $ 868⫾1,519 1,124⫾1,339 1,054⫾1,346 ⬍0.001 *Values are given as mean⫾SD or No. (%).

and guidelines, after they are initially tested and

implemented, is a common limitation of such

meth-ods aimed at standardizing medical practices.

_The

use of automated protocols using prompts and

re-quired inputs from bedside health-care providers

represents one strategy for ensuring more uniform

levels of protocol compliance over time.

Another important limitation of this study is that

we only had experienced RCPs involved in

perform-ing patient assessments. This was purposely done to

obtain support for the performance of this study

from the admitting physicians. Therefore, these

re-sults may have differed if we had used

less-experi-enced individuals in this capacity. Additionally, the

numbers of treatments other than nebulizers and

metered-dose inhalers were small, limiting the

power of our results for those interventions. Lastly,

RCPs in the control arm of this study were not

impeded from communicating their clinical opinions

to the treating physicians. We did not track this

communication, nor did we monitor how often

phy-sicians deviated from the opinions of the RCPs in

firms B and C. A recent investigation suggests that

this may be an important barrier to the

implemen-tation of RCP-directed guidelines.

In summary, the use of RCP-directed treatment

protocols decreased the overall use of respiratory

care and decreased respiratory care charges without

resulting in any detrimental clinical outcomes.

Pro-tocol-guided treatment represents an effective

strat-egy for standardizing medical practices and focusing

the efforts of health-care workers on specific tasks

needing to be performed in a timely manner.

Addi-tionally, protocols can serve as important educational

tools providing clinicians information on the

appro-priateness of various medical practices for specific

disease processes.

_{However, the effective use of}

protocol-directed therapies requires a dedicated

ef-fort on the part of the institution to ensure its

success. This implies that such protocols are

main-tained and updated on a regular basis to optimize

their clinical applicability, have adequate support

staff in place to perform the necessary treatments,

and do not impair or interfere with the ability of

clinicians to alter practices based on their own

experiences.

References

1 Kollef MH. Outcomes research as a tool for defining the role of respiratory care practitioners in the ICU setting. New Horiz 1998; 6:91–98

2 Zibrak JD, Rossetti P, Wood E. Effect of reductions in respiratory therapy on patient outcomes. N Engl J Med 1986; 315:292–295

3 Celli BR, Rodriguez KS, Snider GL. A controlled trial of intermittent positive pressure breathing, incentive spirome-try, and deep breathing exercises in preventing pulmonary complications after abdominal surgery. Am Rev Respir Dis 1984; 130:12–15

4 Johnson D, Kelm C, To T, et al. Post-operative physical therapy after coronary artery bypass surgery. Am J Respir Crit Care Med 1995; 152:953–958

5 Hall JC, Tarala RA, Tapper J, et al. Prevention of respiratory complications after abdominal surgery: a randomized clinical trial. BMJ 1996; 312:148 –152

6 Alexander E, Weingarten S, Mohsenifar Z. Clinical strategies to reduce utilization of chest physiotherapy without compro-mising patient care. Chest 1996; 110:430 – 432

7 Dhand R, Jubran A, Tobin MJ. Bronchodilation delivery by metered-dose inhaler in ventilation-supported patients. Am J Respir Crit Care Med 1995; 151:1827–1833

8 Idris AH, McDermott MF, Raucci JC, et al. Emergency department treatment of severe asthma: metered-dose in-haler plus holding chamber is equivalent in effectiveness to nebulizer. Chest 1993; 103:665– 672

9 Colacone A, Afilalo M, Wolkove N, et al. A comparison of albuterol administered by metered dose inhaler (and holding chamber) or wet nebulizer in acute asthma. Chest 1993; 104:835– 841

10 Pilon CS, Leathley M, London R, et al. Practice guideline for arterial blood gas measurement in the intensive care unit decreases numbers and increases appropriateness of tests. Crit Care Med 1997; 25:1308 –1313

11 Kirby EG, Durbin CG Jr. Establishment of a respiratory assessment team is associated with decreased mortality in patients re-admitted to the ICU. Respir Care 1996; 41:903– 907

12 Stoller JK, Skibinski CI, Giles DK, et al. Physician-ordered respiratory care vs physician-ordered use of a respiratory therapy consult service: results of a prospective observational study. Chest 1996; 110:422– 429

13 Stoller JK, Mascha EJ, Kester L, et al. Randomized controlled trial of physician-directed versus respiratory therapy consult service-directed respiratory care to adult non-ICU inpatients. Am J Respir Crit Care Med 1998; 158:1068 –1075

14 Connors AF, Hammon WE, Martin RJ, et al. Chest physical Figure4. Box plots for respiratory care charges in patients with

and without discordant respiratory care orders. Boxes represent 25th to 75th percentiles, with 50th percentile (solid line) and mean (broken line) values shown within the boxes. The 10th and 90th percentiles are shown as capped bars, and closed circles mark the 5th and 95th percentiles.

therapy: the immediate effect on oxygenation in acutely ill patients. Chest 1980; 78:559 –564

15 Kollef MH. Prolonged use of ventilator circuits and ventila-tor-associated pneumonia: a model for identifying the optimal clinical practice. Chest 1998; 113:267–269

16 Stoller JK, Haney D, Burkhart J, et al. Physician-ordered respiratory care versus physician-ordered use of a respiratory therapy consult service: early experience at the Cleveland Clinic Foundation. Respir Care 1993; 38:1143–1154 17 Hilling L, Bakow E, Fink J, et al. Postural drainage therapy:

American Association for Respiratory Care practice guideline. Respir Care 1991; 36:1418 –1426

18 American Association for Respiratory Care. AARC clinical practice guideline: assessing response to bronchodilator ther-apy at point of care. Respir Care 1995; 40:1300 –1307 19 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a

severity of disease classification system. Crit Care Med 1985; 13:818 – 829

20 Menzies R, Gibbons W, Goldberg P. Determinants of wean-ing and survival among patients with COPD who require mechanical ventilation for acute respiratory failure. Chest 1989; 95:398 – 405

21 Pingleton SK, Fagon JY, Leeper KV Jr. Patient selection for clinical investigation of ventilator-associated pneumonia: cri-teria for evaluating diagnostic techniques. Chest 1992; 102: 553S–556S

22 SAS Institute. SAS/STAT User’s Guide (vol 2). Cary, NC: SAS Institute, 1990; 1071–1126

23 Concato J, Feinstein AR, Holford TR. The risk of determin-ing risk with multivariable models. Ann Intern Med 1993; 118:201–210

24 Wood G, Macleod B, Moffatt S. Weaning from mechanical ventilation: physician-directed versus a respiratory-therapist-directed protocol. Respir Care 1995; 40:219 –224

25 Kollef MH, Shapiro SD, Silver P, et al. A randomized, controlled trial of protocol-directed versus physician-directed weaning from mechanical ventilation. Crit Care Med 1997; 25:567–574

26 Joiner GA, Salisbury D, Bollin GE. Utilizing quality assurance as a tool for reducing the risk of nosocomial ventilator-associated pneumonia. Am J Med Qual 1996; 11:100 –103 27 Kelleghan SI, Salemi C, Padilla S, et al. An effective

contin-uous quality improvement approach to the prevention of ventilator-associated pneumonia. Am J Infect Control 1993; 21:322–330

28 Goldberg R, Chan L, Haley P, et al. Critical pathway for the emergency management of acute asthma: effect on resource utilization. Ann Emerg Med 1998; 31:562–567

29 Ford RM, Phillips-Clar JE, Burns DM. Implementing ther-apist-driven protocols. Respir Care Clin North Am 1996; 2:51–76

30 Peterson WL, Cook DJ. Using a practice guideline for safely shortening hospital stay for upper gastrointestinal tract hem-orrhage. JAMA 1997; 278:2186 –2187

31 Grimshaw JM, Hutchinson A. Clinical practice guidelines: do they enhance value for money in health care? Br Med Bull 1995; 51:927–940

32 Clemmer TP, Spuhler VJ. Developing and gaining acceptance for patient care protocols. New Horiz 1998; 6:12–19 33 Horst HM, Mouro D, Hall-Jenssens RA, et al. Decrease in

ventilation time with a standardized weaning process. Arch Surg 1998; 133:483– 488

34 Clemmer TP, Spuhler VJ, Berwick DM, et al. Cooperation: the foundation of improvement. Ann Intern Med 1998; 128:1004 –1009

35 Pearson SD, Goulart-Fisher D, Lee TH. Critical pathway as a strategy for improving care: problems and potential. Ann Intern Med 1995; 123:941–948

36 Lobach DF. Electronically distributed, computer-generated, individual feedback enhances the use of a computerized practice guideline. Proc AMIA Annu Fall Symp 1996; 493– 497

37 Nilasena DS, Lincoln MJ. A computer-generated reminder system improves physician compliance with diabetes preven-tion care guidelines. Proc Annu Symp Comput Appl Med Care 1995; 640 – 645

38 Ely EW, Bennett PA, Bowton DL, et al. Large scale imple-mentation of a respiratory therapist-driven protocol for ven-tilator weaning. Am J Respir Crit Care Med 1999; 159:439 – 446

39 Thorens JB, Kaelin RM Jolliet P, et al. Influence of the quality of nursing on the duration of weaning from mechan-ical ventilation in patients with chronic obstructive pulmonary disease. Crit Care Med 1995; 23:1807–1815