The use of structured, complaint-specific patient encounter forms in the emergency department

ORIGINAL C O N T R I B U T I O N patient satisfaction

The Use of Structured, Complaint-Specific Patient

Encounter Forms in the Emergency Department

From the Department of Emergency Medicine, Va~derbiIt University, Nashville, Tennessee;* and the Division of Emergency Medicine t and Department of Pediatrics,~: University of Rochester School of Medicine, Rochester, New York.

Received for publication June 5, 1992. Revision received October 19, 1992. Accepted for publication November 25, 1992.

This study was supported by an Innovations in Patient Care grant from Strong Memorial Hospital.

Keith Wrenn, MD* Lance Rodewald, MD ~r

Eileen Lumb, RN t

Corey Slovis, MD*

Study objective: To assess the effect of preprinted, structured, complaint-specific patient encounter forms on documentation, use of testing, and treatment compared with free-text record keeping.

Design: Nonrandomized case-control trial.

Setting: University-affiliated, tertiary referral hospital emergency department.

Methods: The records of all patients with lacerations, pharyngitis, asthma, or isolated closed-head injury during an eight- month period were reviewed.

Intervention: Use of structured complaint-specific patient encounter forms versus traditional free-text record keeping.

Main outcome measure: The null hypothesis was that there would be no differences in documentation, test use, or practice when the structured forms were used compared with free-text record keeping.

Results: Differences in documentation that favored the use of the structured forms for all four problems studied were seen consistently. Not only was documentation improved, but test use also was affected in a way that decreased use. In addition, in certain areas (eg, treatment of pharyngitis), clinical practice also was changed.

Conclusion: Structured, problem-specific ED records improve documentation and affect both resource use and clinical practice. These forms may be useful for improving communication and reimbursement as well as for medicolegal documentation. They provide a method for standardized quality assurance review and clinical data abstraction. Finally, they provide a method for active dissemination of clinical standards.

[Wrenn K, Rodewald L, Lumb E, SIovis C: The use of structured, complaint-specific patient encounter forms in the emergency department. Ann Emerg Med May 1993;22:805-812.]

MAY 1993 22:5 ANNALS OF EMERGENCY MEDICINE 9 0 5 / 5 5

ENCOUNTER FORMS Wrenn et al

INTRODUCTION

There has been a movement over recent years to establish flexible and rational clinical guidelines and standards on a national level. 1,2 These guidelines generally have taken the form of recommendations about the usefulness of procedures or the management of common conditions. Unfortunately, dissemination of these standards by "diffu- sion" has not been shown consistently to improve practitioners' behavior. 3-7

In addition to developing standards, there is a need to assess whether such standards are upheld. Clinical quality assurance methods are the major part of that process, but they often entail labor-intensive, retrospective reviews of a sample of medical records. These reviews are dependent on provider documentation and the specific expertise or interests of the reviewer, s Although such reviews can outline general areas that will need improvement in the future, potential errors occur after the fact, often necessi- tating the time-intensive efforts of calling back patients, arranging follow-up, and educating the person who made the error.

In the emergency department, quality assurance .issues take on special importance because care is given for a wide variety of conditions by multiple practitioners from a variety of specialties in a time-intense environment. Apparent errors or quality issues often result from a simple lack of adequate documentation. At other times, quality issues reflect a basic lack of knowledge about what is required to diagnose and properly treat a particular condition.

Many common chief complaints in the ED lend themselves to a directed evaluation that is comprehensive and efficient. We developed patient-encounter forms that are complaint-specific and structured. They were designed to include what the emergency physicians and nurses thought were the key items in the history and physical examination as well as guidelines for the use of laboratory

Table 1. Age and sex of patients with one of the study chief complaints

Chief Use of Use of Complaint Quicksheet Free Text P

Age (y r ) Laceration 23.2 + 18.7 24.9 + 17.9 NS (mean +SD) Pharyngitis 26.9+10.5 18.6+11,4 < .0000

Asthma 14.7 +16.3 16.9_+17.8 NS Olased-head Injury 13.6_+16.4 14.7_+10,9 NS

% Malesex Laceration 68 69 NS Pharyngitis, 34 34 NS Asthma 56 51 NS Closed-head Injury 59 60 NS

tests and radiographs. We called these forms "quicksheets" because they involved checking off items with minimal use of free text. We undertook a prospective evaluation of these forms to assess whether they changed documentation of clinical variables, the use of certain tests, or the manner in which these problems were treated compared with the use of traditional free-text record keeping. We were interested specifically in studying relatively high- volume problems that crossed age and service boundaries. We chose to study asthma, pharyngitis, laceration, and isolated closed-head injury because they occurred in all age groups (Table i) and involved the participation of house officers from the departments of medicine, pediatrics, and surgery.

MATERIALS AND METHODS

Over eight months, every ED chart with a diagnosis of asthma, pharyngitis, laceration, or closed-head injury was reviewed prospectively, and the data elements listed in Tables 2 and 3 were abstracted from the records. The investigators were not involved directly in the care of the patients.

The setting for our study was an urban tertiary referral center and teaching ED that has approximately 60,000 visits annually. Structured medical records (qutcksheets) were developed by the authors in conjunction with faculty members who had expertise in the problem covered by the quicksheet. Effort was directed toward creating documents that reflected a consensus view of the necessary data to be collected for each patient complaint. Consensus was arrived at when all of the six nurses and physicians included in creating the documents agreed to the inclusion or exclusion of an item. Textbooks, position papers, Centers for Disease Control recommendations, specialists in appropriate fields, and published guidelines from specialty colleges such as the American College of Physicians and American College of Emergency Physicians were used in developing the quicksheets.

Table 2. Documentation of variables

% % Use of Use of Odds Ratio

Variable Quicksheet FreeText (99%CI) P

Case discussed with private physician 5 5 1.06 (0.7 - 1.7) NS

Follow-up arranged 98 94 2.5 (1,5 - 4.6) .0000 Drug prescription documented 80 73 1.4 (1.0-2.0) .007

5 6 / 80 6 ANNALS OF EMERGENCY MEDICINE 22:5 MAY 1993

ENCOUNTER FORMS Wrenn et aI

Table 3. Data items reviewed for each specific problem

Variable

% Use of

Quicksheet

% Use of

Free Text Odds Ratio

(99% CI) P

Laceration Mechanism of injury documented 99.8 Length of laceration documented 99 Distal function before sutures documented 99 Sedation use documented 80 Type of anesthesia documented 97 No. of sutures documented 98 Radiograph taken 10 Functional outcome after sutures documented 97 Tetanus immunization documented 99.6 Child abuse indicators documented 98.4 Pediatric ED attending consulted 72 Plastic surgery consultation obtained 1

Closed-Head Injury Mechanism documented 100 Loss of consciousness documented 100 Duration of loss of consciousness documented 93 Glasgow Coma Scale documented 97 Presence or absence of periorbital ecchymosis documented 98 Presence or absence of hemotympanum documented 97 Presence or absence of cerebral spinal fluid leak documented 97 Cervical-spine examination documented 96 Cervical-spine radiograph done 19 Head computed tomography scan done 5 Skull film done 12 Discharged to care of adult documented 92 Head injury instruction sheet documented or given 95 Neurosurgical consultation obtained 6

Pharyngitis Temperature documented by house officer 99 Presence or absence of nodes documented 100 Presence or absence of tonsillar exudate documented 99 Throat culture performed 48 Rapid streptococcal assay performed 12 Any radiograph done 6 Diagnosis of Streptococcus versus of other 68 Oral penicillin given 4 Intramuscular benzathine penicillin given 33 Prescription length documented 44 Drug allergy documented 99

Asthma Precipitant documented 98 Home medications documented 99.6 Prior admission documented 96 Pulsus paradoxus documented 31 Pulmonary function tests documented 29 Theophylline level documented 43 Chest radiograph performed 19 Oximetry performed 66 Arterial blood gases performed 1 IV theophylline given in ED 16 IV steroid given-in ED 25 Theophylline given on discharge 51 Steroid given on discharge 71 6-Agonist given on discharge 98

*Unable to be calculated by crass-product ratio; number is too large or too small or denomination is O.

98.5 84 71

5 76 71 19 16 95 94.4 49 11

100 98 84 17 25 71 56 76 22 14 7

87 86 7

61 96 98 60 25

5 59

8 8

45 99.6

90 94 59 3

20 54 36 54 11 19 28 49 61 95.3

7.8 (0.8 - *) 20.1 (6.3 - 64.7) 30.2 (11.1 -82.1) 71.0 (12 - *) 10.5 (4.6 - 23.6) 19.9 (8 - 50) 0.45 (0.3 - 0.8)

163 (74 - 354) 13.1 (2.3- *) 0.28 (0 - 1,5) 2.7 (1.2 - 6) 0.08 (0.03- 0.28)

*

2.3 (0.4 - *) 176 (41 - *) 165 (31 - *)

14.6 (3.5 - *) 27.4 (6.6 - *)

8.2 (2.3 - *) 0.87 (0.4-1.8) 0.29 (0.09 - 0.96) 1.9 (0.7 - 5.1) 1.3 (0.5 - 3.5) 3.3(1-11) 0.87 (0.3 - 2.7)

44.0 (13 - *)

1.8 (0,3 - *) 0,62 (0.4 - 1) 0.40 (0.2 - 0.7) 1.1 (0.4-2.8) 1.4 (0.9 - 2,3) 0.42 (0.2 - 1.1) 5.3 (2.7 - 10.3) 0.93 (0.6 - 1.5) 0.5 (0.8 - *)

5.1 (1.6 - 16) 11.5 (1.3 - *) 16.9(7-41) 14.3 (6.8 - 30)

1.6 (I - 2.5) 0.6 (0.4 - 1.1) 0.4(0.2-0.7) 1.7 (1.1-2,5) o.1 ( * -o .5) 0.8 (0.5 - 1.4) o.8 (0.5-1.5) 1.1 (0.7 - 1.7) 1.5 (1 - 2.5) 2.6 (0.7 - *)

NS <.O000 <.0000 <.0000 <.0000 <.0000

.0002 <.0000 <.0000

NS .002

<.0000

NS NS NS

< .0000 < .0000 < .0000 < .0000 < .0000

NS .008 NS NS

.01 NS

< .0000 .0005

NS .006 .0001 NS NS

.02 < .0000

N8 NS

.0002

.003 < .0000 < ,0000

.013

.02 < .0000

.002 < ,0000

NS N8 NS

.017 NS

MAY 1993 22:5 ANNALS OF EM£RGENCY MEDICINE 8 0 7 / 5 7

E N C O U N T E R F O R M S

Wren~ et aZ

All quicksheets were reviewed by the hospital forms committee and authorized to be placed in patients' per- manent medical records as a substitute for the standard free-text form. At this level of review, a large number of nurses and physicians from several departments had suggestions, and some changes occurred in content and format. The study was approved by the research subjects review board; informed consent from the patient was not required for entrance into the study

Attending ED faculty were instructed at facuhy meet- ings in the proper use of the quicksheets. ED housestaff, consisting of trainees at postgraduate levels 1 through 3 in surgery, medicine, and pediatrics, were instructed at morning teaching conferences for a period of five to ten minutes on the use of the quicksheets. All subjects knew that a quicksheet study was in progress, but they did not know the hypotheses of the study

Twenty-five physicians who used the quicksheets (64% of the house officers staffing the ED during the study period) returned a questionnaire designed to evaluate their atti- tudes toward these forms. Two had not used the quicksheets at all, and the others all had experience with both quicksheets and free-text records. The questionnaire was answered anonymously and involved the use of a Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree) to assess statements about the quicksheets.

The decision to use the quicksheet or free-text form for an individual patient was made by the housestaff or faculty member caring for the patient; study group membership was determined by whether a quicksheet was used (quicksheet group) or was not used (free-text group). All comparisons were made between these two groups. Determinations of the patients' chief complaints were made by the on-duty triage nurse at the time of presentation and recorded in the medical chart on a page separate from the quicksheet or free-text forms. These determinations defined membership in the problem- specific study groups.

The size of the study groups had been determined previously and was based on a desired ability to detect clinically meaningful differences in rates of documentation of key clinical features of each of the four medical problems. The variables to be studied were chosen before initiation of the study Rates were modeled as binomial variables; the tables of Fleiss 9 were used to determine sample sizes. Enrollment into the eight study subgroups continued until all of the desired sample sizes were exceeded, at which point enrollment ceased. The laceration quicksheet subgroup, for example, filled before the laceration free-text subgroup. We therefore continued

to add laceration quicksheet subjects until the laceration free-text subgroup also had filled, at which time we stopped enrolling patients in either laceration subgroup.

Medical charts of patients with one of the four chief complaints being studied were abstracted by a project nurse who did not provide care for the patients. Chart abstraction was done using a prospectively derived problem-specific data collection form. The same chart abstraction was used regardless of whether the patient was in the quicksheet or the free-text group. Tables 2 and 3 show all of the comparisons that were made and completely define the content of the data collection forms.

Comparisons were made only within the problem- specific groups, with the exception of Table 2 variables, which pertained to all four problems. All comparisons of rates were made using the Z 2 statistic, correcting for multiple within-group comparisons using the Bonferroni correction. 9 The a value of .05 was divided by the number of comparisons in that subgroup to determine the accept- able P value (eg, there were 12 laceration variables, so we used a P cutoff of .05/12, or .004). Odds ratios and 99% confidence intervals were calculated for all comparisons.

RESULTS

Between October 1990 and May 1991, 2,405 charts were reviewed. These included 779 laceration records (513 quicksheets and 266 free text), 526 pharyngitis records (274 quicksheets and 252 free text), 730 asthma records (233 quicksheets and 497 free text), and 370 closed-head injury records (t09 quicksheets and 261 free text). Baseline demographic data among the groups are recorded in Table 1, and no differences between these groups are noted with the exception of age in the pharyngitis patients (a difference that does not appear to be clinically important).

There was only one instance in which documentation appeared to suffer when the quicksheet was used (documentation of a theophylline level in asthma patients), but statistical significance was not reached for this difference in documentation. In all other cases, documentation was improved by use of the quicksheet compared with free- text record keeping. Variables examined for all forms were documentation of discussion of the case with a private physician, arrangement of follow-up, and complete documentation of prescription information. Documentation of the latter tWO items was improved or tended to improve with use of the quicksheets (Table 2).

Table 3 lists the various data items reviewed prospectively for each specific problem. When Bonferronig correction is applied within groups, there still is statistical

5 8 / 80 8 ANNALS OF EMERGENCY MEDICINE 22:5 MAY 1993

ENCOUNTER FORMS Wrenn et aZ

significance to several of the differences in documentation seen on quicksheets versus free-text records.

Laceration records Differences were found in documentation of length of laceration, distal function before sutures, use of sedation, type of anesthesia used, number of sutures, whether a radiograph was taken, functional outcome after suturing, documentation of the status of tetanus immunization, ED pediatric attending consultation, and plastic surgery consultation. In all these instances, quicksheet documentation was more complete.

Closed-head injury records Statistically and clinically significant differences were found for documentation of the Glasgow Coma Scale, presence or absence of periorbital ecchymosis, presence or absence of hemotympanum, presence or absence of a cerebrospinal fluid leak, and documentation of cervical-spine examination. Again, quicksheet documentation was more complete.

Pharyngitis records Documentation of the temperature by the house officer and presence or absence of anterior cervical nodes were better documented in the quicksheet group. Fewer rapid streptococcal tests were performed in the quicksheet group, and intramuscular penicillin was administered more often in the quicksheet group.

Asthma records Statistically significant differences were seen for the variables documentation of precipitant, home medications, prior admissions, and pulsus paradoxus. In addition, there were differences in the rates of performance of a chest radiograph, performance of oximetry, and performance of an arterial blood gas (P < .05/14, or < .0035). There was a trend toward decreased documentation of theophylline levels in the quickshee[ group and increased use of pulmonary function tests.

Cost Some of the variables represented items that add to the cost of an ED visit. Fewer radiographs were taken and fewer plastic surgery consultations were obtained for lacerations when the quicksheet was used. Although there were more frequent consultations with pediatric ED attending physicians when quicksheets were used, this did not add to the cost of a visit. There was a trend toward ordering fewer computed tomography scans of the head in the quicksheet group for closed-head trauma. There were, however, no differences in the numbers of cervical-spree or skull radiographs obtained or the numbers of neurosurgical consultations. In asthma patients, fewer chest radiographs and arterial blood gases were performed in the quicksheet group, but pulse oximetry was used more often. There was a trend toward increased documentation of puhnonary function tests but no differences in the types of medications used. For pharyngitis patients, radiographs of any kind (soft-tissue neck or chest radio-

graph) were used equally. Rapid streptococcal antigen assays were performed less often, and there was a trend toward decreased use of throat cultures in the quicksheet group, even though the discharge diagnosis of "streptococcal pharyngitis" was used equally often.

Only for asthma were there large enough numbers of admitted patients to assess differences in final disposition from the ED. There was no difference in the percentage of patients admitted for asthma in the qnicksheet group (5%) versus the free-text group (6%).

The response of house officers from the departments of medicine, pediatrics, and surgery generally were favorable when responding to statements about the quicksheets. There was strong agreement or agreement in 100% of house officers that the forms were easy to use, and 78% agreed or strongly agreed that the quicksheets made evaluation of the patient faster. There was agreement by 61% of house officers that the forms contributed to their knowledge base. Seventy-eight percent of house officers disagreed with the idea that the forms limited their thinking process, and 65% disagreed with the idea that the forms limited the differential diagnosis. Seventy-four percent of house officers thought that the forms made them include information they otherwise would not have included, and only 30% thought that there were significant variables in the history and physical examination that were missing from the forms. Thirty-nine percent felt more comfortable medicolegally when they used the forms. Finally, 91% thought that these forms should be used in the future, and 78% thought that more forms should be developed.

DISCUSSION

The medical record serves many functions and means different things to different groups of health care providers. It is the primary way health professionals communicate with one another, allows information to be preserved, and facilitates continuity of care. lo It is the source document when questions of medical negligence or malpractice arise, lo Under the new relative value resource based system reimbursement scheme, documentation affects physician reimbursement. It often is the only information used in quality assurance activities. Finally, the record can serve as a source of data for clinical research as well as research into "process of care" for purposes of planning strategies to impact on the overall cost of health care delivery locally and nationally. 1.11

The quicksheets provided concise, complete information about the ED visit for specific chief complaints. They

MAY 1993 22:5 ANNALS OF EMERGENOY MEDJCINE 8 O 9 / 5 9


facilitated data recording and communication by improving documentation. Better documentation should make defense of malpractice claims easier. There are fewer problems of legibility when they are used, and they decrease the time it takes to do paperwork.

The implications for research from these sheets are multiple. Abstracting data retrospectively is time consuming because of the dual problems of poor legibility and nonstandardized write-ups with free-text records. Therefore, data abstraction can be extremely difficult. Abstraction is simplified dramatically with these sheets because data are recorded in the same fashion for every patient. Similarly, quality assurance review should be simpler and more objective. Because the sheets may be placed on a menu-driven computer screen using light-pen input, data abstraction and quality assurance may occur simultaneously with data entry rather than retrospectively

There is substantial literature on whether and how guidelines can be used to improve clinical practice and decrease practice variations. 3-r, 11 In an academic setting, structured patient encounter forms appear to be a way to disseminate information and teach clinical standards actively rather than passively, thereby ensuring positive changes in clinical behavior. Quicksheets disseminate this information during "teachable moments," those moments where individual clinicians are providing care to a patient with the specific problems to which the standards apply r They also serve as continuous reminders, an important point in terms of maintaining changes in behavior. 12 This is important because passive dissemination of clinical standards by publication (eg, in journals) has been disappointing. 3-z These forms might decrease variations in care across geographic, economic, and cultural boundaries. 10-14 These issues are especially important in the time-intensive environment of the ED where physicians may be from many specialties with varying expertise. "Outcome" is as important in the ED as it is in any other setting but may be more process dependent.15,10 Decision making in the ED often is process intensive because the providers caring for a patient often have no prior relationship with the patient.

The complaint-specific, preprinted, structured patient encounter forms improved documentation, but they did not increase the cost of delivering care. In fact, cost savings were evident in several areas. Cost savings possibly resulted from the guidelines for ordering of tests. In particular, the use of arterial blood gases and chest radiographs- in asthma patients, compute d tomography scans of the brain in closed-head injury patients, and throat cultures and rapid streptococcal antigen tests in pharyngitis patients were

i

decreased when quicksheets were used. The sensitivity of rapid streptococcal antigen tests has been shown to be little better than flipping a coin in predicting streptococcal pharyngitis. 1 z

Clinical practice was changed in other ways. For patients with lacerations (usually seen initially by a surgical house officer), pediatric ED attending physicians were involved more often when the quicksheets were used, probably because of a structured reminder about child abuse indicators and discussions with the pediatric attending. Involvement of the pediatric ED attending physicians does not increase costs in our institution because only one bill for professional services is sent even if more than one ED attending sees the patient. In addition, the length of lacerations and the number of sutures used were recorded more reliably on the quicksheets, which should lead to improvements in reimbursement. Among patients with pharyngitis, intramuscular benzathine penicillin was given more often to treat presumed streptococcal pharyngitis when the quicksheets were used. Intramuscular benzathine penicillin has been shown to be well tolerated and to have better cure rates than oral penicillin in streptococcal pharyngitis. 18,19

A recent prospective trial also showed that preformatted charts for the gynecologic complaints of abdominal pain, bleeding, or vaginal discharge improved documentation in the ED. 2o Our study showed similar effects and adds to the literature by increasing the number of problems studied.

Good medical records do not necessarily reflect good medical performance. Some have suggested a correlation between the quality of the record and the quality of care. lo,~ ~ Others, in retrospective fashion, have not been able to find a relationship between the quantity or quality of recorded data and outcomes. 15 Retrospective review of a record's quality may not correlate with outcome because there often is a difference between what is recorded and what is actually done (the problem of "underreporting" in the interest of time). 21 The quicksheets used in this study, however, help ensure prospectively that data are not only recorded but also collected (assuming the professional integrity of the physician filling out the quicksheet). It seems intuitive that decision making would be improved if the collection of critical variables in diagnosis is enforced by reminders in the form of checklists.

This study was not a randomized clinical trial. Because group assignment was made at the discretion of the physicians whose documentation was being studied, the potential existed for selection bias. This bias precluded us from crediting the changes in practice solely to the use of the quicksheets. Several types of selection bias were possible.

6 0 / 8 1 0 ANNALS OF EMERGENCY MEDICINE 22:5 MAY 1993


First, the choice to use free text or quicksheets may have been determined by severity of illness (ie, more severe illness was documented more often by free-text recording). That fewer plastic surgery consultations or radiographs were obtained in the quicksheet laceration group probably reflects the facts that lacerations requiring such consultations or radiographs were more complicated and that free-text recording made more sense for these patients. Another possibility is that the initial provider did not document but instead secondarily triaged the patient directly to the consultant. That there was no difference in the rate of consultation with neurosurgery for closed-head injuries, regardless of the record-keeping method chosen, argues against this type of bias in the other quicksheets. In addition, for asthma patients there were no differences in admission rate, use of steroids or IV theophylline in the ED, or use of steroids on discharge. There were no differences in the numbers of radiographs performed or the diagnosis on discharge for pharyngitis patients.

A second potential bias was that physicians who liked the quicksheets tended to use them in lieu of free-text recording, and, conversely, those preferring free-text recording preferentially avoided the quicksheets. If this bias was in effect, and we believe that this was likely, one would expect to see smaller differences in documentation. This would strengthen our observation of large differences in documentation. It also was possible that physicians choosing to use the quicksheet would have documented more thoroughly on the free-text form, but this seems unlikely.

A third potential bias was that more compulsive or more meticulous record keepers would tend to use the form rather than free text. This seems unlikely given the almost uniformly positive responses of the house officers, who represented three different departments.

External validity should not be a problem if these results are generalized only to other urban teaching EDs, where housestaff physicians and medical students perform the vast majority of the medical chart documentation.

We encountered considerable institutional prejudice against the use of the quicksheets. Several attending physicians thought that these forms might limit the creative thinking of house officers. We believe that thinking is separate from what appears on the chart, and there is evidence to support this claim in papers showing a disparity between what appears on the chart and the quality of care delivered. 15

Before the study, several house officers expressed concern about using the forms based on their perspective of limited freedom of expression. After using the forms,

however, the house officers did not repeat this complaint. The concerns about limiting freedom of expression and thinking have to be balanced against ensuring that adequate information is collected and documented, Minimal clinical standards must be met, and there is no evidence that allowing freedom in the form of free-text record keeping encourages meeting standards. There is, however, considerable evidence that documentation and incorporation of standards have not occurred routinely through passive methods of dissemination, including didactic teaching and publication.>ra 1 These forms do not have to limit expression. Other items can be added to the form in free text, or a supplemental free-text note on standard hospital "progress note" paper can be added on any patient.

CONCLUSION

We believe that many problems and complaints common- ly seen in an ED lend themselves to preprinted, structured data collection and clinical encounter forms. These forms improve documentation and are useful for communication, medicolegal defense, billing, and research. They also may serve to disseminate predetermined explicit, well- defined minimal clinical standards in a way that teaches and changes clinical behavior. They are, therefore, power- ful tools that must be designed carefully to diminate the possibility of a negative impact. They represent a proac- rive, prospective approach to quality assurance and are valuable, not only for populations but also for individual patients and practitioners. Because they can be-computer generated, they may be cost efficient, flexible, and able to be updated. 22 They permit integrated documentation of nursing and physician functions in collaborative practice plans that would eliminate redundancy and potentially enhance outcome. Further research into the effect of such forms is needed, particularly in the area of variations in care.

REFERENCES 1. Kapp MB: "Cookbook" rnedicine--A legal perspective. ArchlnternMed1990;150:496-500.

2. Brook RH: Quality of care: Do we care? Ann Intern Meal 1991;115:486-490.

3. Lomas J, Haynes RB: A taxonomy and critical review of tested strategies for the application of clinical practice recommendations: From "official" to "individual" clinical policy. Am JPrevMed 1988;4 {suppl):77-84.

4. Lornas J, Anderson GM, Darnnick-Pierre K, et el: Do practice guidelines guide practice? The effect of a consensus statement on the practice of physicians./V Eng/J Med1989;321:1306-1311.

5. Lama J: Holding back the tide of caesareans: Publishing recommendations is not enough. Br Med J 1988;217:569-570.

6. Kanouse DE, Jacoby I: When does information change practitioners behavior? IntJ Tech Assess Health Care 1988;217:569-570.

MAY 1993 22:5 ANNALS OF EMERGENCY MEDICINE 8 1 1 / 6 1

ENCOUNTER FORMS Wrcnn et al

7. Soumerai SB, Avem J: Principles of educational outreach ("academic detailing") to improve clinical decision making. JAMA 19£0;263:549-556.

8. Goldman RL: The reliability of peer assessments of quality care. JAMA 1992;267:£58-960.

9. Fleiss JL: Statistical Methods forRates and Proportions, ed 2. New York, John Wiley & Sons, 1981.

10. Lyons T, Payne 8: The relationship of physicians' medical recording performance to their medical care performance. Med Care 1974;12:463-469.

11. Zuckerman AE, Starfield B, Hochreiter C, et ah Validating the content of pediatric outpatient medical records by means of tape-recording doctor-patient encounters. Pediatrics 1975;56:407- 411.

12. Tierney WM, Miller ME, McDonald C J: The effect on test ordering of informing physicians of the changes for outpatient diagnostic tests. N Eng/J Med 1990;322:1499q 504.

t3. Wennberg JE, Blowers L, Parker R, et el: Changes in tonsillectomy rates associated with feedback and review. Pediatrics 1977;59:821-826.

14. Wennberg JE: Dealing with medical practice variations: A proposal for action. Health Aft 1984;3:6-32.

15. Fessel W J, Van Brunt EE: Assessing quality of care from the medical record. N Engl J IVied 1 £72;285:I 34-138.

16. Kessner DM: Quality assessment and assurance. Early signs of cognitive dissonance. N Engl J Med 1978;298:381-386.

17. Wegner DL, WEe DL, Schrantz RD: Insensitivity of rapid antigen detection methods and single blood agar plate culture for diagnosing streptococcal pharyngitis. JAMA 1992;267:695-697.

18. Breese BB, Disney FA: Penicillin in the treatment of streptococcal infections: A comparison of effectiveness of five different oral and one parenteral form. N Eng/J Med1958;259:57-62.

19. Kaplan EL: Benzathine penicillin G for treatment of group A streptococcal pharyngitis: A reappraisal in 1985. Pediatr Infect Dis 1985;4:592-596.

20. Humphreys T, Shofer FS, Jacobsen S, et ah Preformatted charts improve documentation in the emergency department. Ann Emerg Med 1992;21:534-540.

21. Duggan AK, Starfield B, DeAngelis C: Structured encounters form: The impact an provider performance and recording of well-child care. Pediatrics 1990;85:104q 13.

22. Karcz A, Holbrook J: The Massachusetts emergency medicine risk management program. ORB 1991;17:287-292.

Copies of the preprinted quicksheets used in this study are available from Dr Wrenn. Please enclose a stamped, self-addressed envelope.

Address for reprints:

Keith Wrenn, MD

Department of Emergency Medicine

Room 1368

VanderbiIt University Medical Center

Nashville, Tennessee 37232

6 2 / 8 1 2 ANNALS OF EMERGENCY MEDICINE 22:5 MAY 1993

Documents

The use of structured, complaint-specific patient encounter forms in the emergency department