|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Being unaware of previously occurring in-hospital and perioperative events can lead to perioperative disaster in a subsequent surgical procedure. For example, an unanticipated difficult airway or allergic reaction that occurred earlier but is not known can (and has) contributed to disasters in subsequent anesthesia care. Avoiding such disasters would seem to be a solvable problem in an age of records, paper or otherwise. Yet anyone who has practiced in a large modern-day American hospital (university or community) knows that past charts are often not available. Even when the patient is seen in a surgeon's office or perioperative clinic, the transfer of records and consent forms can be a weak link in the chain of communication. Enter the information age and the promise of making the chain perfect.
Because we currently practice in an environment called the "Information Age," how is it possible that we still have trouble obtaining the information we need? As our health care institutions switch to electronically stored patient records, theoretically this problem should vanish. However, these institutions and a variety of vendors that now offer programs for automating the preoperative assessment frequently fail to provide operational solutions. These "turnkey" products are often isolated implementations of paper records. Such preoperative systems have not achieved wide acceptance, despite their introduction more than 3 decades ago.[275] [276] However, planned with the overall goals of the preoperative evaluation in mind, information systems can be used successfully to improve patient care and the efficiency of the overall system.
Such a system requires an accurate and complete database storing the patient's current health status and past anesthesia and surgical experiences. The sources for the data may be varied and have been collected at disparate locations and times. The ideal database should include the history of the current surgical problem plus past medical, social, and surgical histories. It should also include review of organ systems, current physical examinations and perioperative plans, and notes on discussions with the patient and significant others. This database can also be used when providing ancillary services such as laboratory tests and consultations. Once the perioperative plan has been formulated, a record can be clearly communicated to the patient, the surgeon, and the anesthesiologist's colleagues. Finally, by establishing an accurate baseline status, the information that has been acquired can be used to initiate quality assurance procedures.
This section provides an introduction to the use of information systems and identifies important issues for the clinician who is considering implementing such a system.
Traditionally, the physician obtained the information he or she needed by talking with the patient, scanning the available paper chart, consulting other physicians, and collecting reports from laboratories or consultants. These data were then recorded in a note. The information was often incomplete at the time the note was written (missing lab data or uncompleted consultation). Thus, the note that may have needed to be updated often wasn't.
To standardize the format and help the physician obtain important information, many facilities use forms that range from simple headings placed on a page to detailed checklists (see Fig. 25-2a and Fig. 25-2b for a checklist our group now sends to surgeons and proceduralists to be completed by the patient prior to seeing an anesthesiologist). As physician contact time has become more limited, ancillary personnel have been recruited to help acquire and assemble the required data. Patients are also being asked to spend more time completing checklists or questionnaires that cover many points in their history. Compared with the current procedure, automation should be able to improve many of these processes. For example, the use of electronic information systems (informatics) can reduce or eliminate the redundant entry of information in a record, a situation that frequently creates discrepancies in that record.
The expense of obtaining these histories increases because we often rely on professionals to obtain the routine history rather than focus on the details that are abnormal or important to the planned care. The process could be less expensive if patients completed the screening process themselves. There are many ways the patient could do this. When patients are given a clear reason for their input and a clear and simple format for answering questions, the patient-completed questionnaire can be an accurate and powerful tool.[6] [276]
One of the first systems used for this purpose was the HealthQuiz. * Developed in 1987, this system consisted of a laptop device that had a simple liquid crystal display and three buttons; the patient pressed a "yes," "no," or "not sure" button to answer health-related questions. The HealthQuiz was a simple system, not unlike many of the small electronic games becoming popular at the time, and it was readily accepted by patients. Its program consisted of a branching algorithm made up of 254 items. The wording of each question was carefully prepared and validated so that consistent answers were given whether the patient was responding to a human or to the computer or was later asked the same question again. The initial questions were standard health history items
Although compact and easy to use, the HealthQuiz required the use of a dedicated unit for producing the questionnaire. In 1993, the rapid expansion of the Internet into health care and even the patient's home prompted the development of a networked version of the system, the HealthQuiz Plus 2, which could be accessed from any computer by means of an Internet browser. This concept has been extended, and now the patient can
Figure 25-15
The HealthQuiz Plus 2 computer device uses the patient's
answers to its questionnaire to produce this summary of the patient's health history
and suggested laboratory tests, annotated with International Classification of Diseases,
ninth edition (ICD-9) codes. BMI, body mass index; BP, blood pressure; BUN, blood
urea nitrogen; CREAT, creatinine; CV, cardiovascular; DOB, date of birth; DOE, dyspnea
on exertion; ECG, electrocardiogram; GI, gastrointestinal; HQASA1-5, Health Quiz
ASA status square; HR, heart rate; HT, height; Hx, history; IVDA, intravenous drug
abuse; PSH, past surgical history; PULM, pulmonary; SOB, shortness of breath; S/P,
status post; WT, weight.
Data are also available at the point of care and are easily transferred to other medical information systems. For example, justification of laboratory tests by the assignment of diagnosis-related codes (ICD-9) is produced for each test suggested, and this information can be given to the laboratory to aid in reimbursement ( Fig. 25-15 and Fig. 25-16 ). Such procedures supersede the old but still often
Figure 25-16
Example of the summary sheet that gives the physician
the rationale for each of the tests suggested for the patient. This sheet is generated
by the HealthQuiz computer device, which uses the patient's answers to the HealthQuiz
questionnaire, plus 480 algorithms programmed into the device, to make its decisions;
it also provides International Classification of Diseases, ninth edition (ICD-9)
codes for each choice. ECG, electrocardiogram.
Although the value of the preoperative evaluation has been discussed, efforts to improve efficiency cause anesthesiologists to consider whether some patients may be assessed just as effectively immediately before surgery. To be successful, this plan should accomplish three key goals:
Optimally the evaluation should be done far enough in advance to allow motivating the patient to adopt healthier behaviors, such as regular physical activity and smoking cessation, and to institute therapies such as aspirin, β-adrenergic receptor blocking drugs, statins, and others that will facilitate better perioperative and long-term outcomes. Tools such as the patient-completed health history, along with accessibility to previous information on perioperative care and other available data, can provide the second leg of the triad with minimal provider resources, identifying patients who need to come to the clinic for additional evaluation. Finally, on-line educational resources, perhaps in association with a brief call from the anesthesiologist, can complete the process of discussing the anesthesia plan with the patient. This plan would allow patients to present on
Physician, hospital, and health system methods for data gathering that include radiology reports, laboratory test reports, and prior operative notes and anesthetic records are now becoming more widespread. They may use not only standard computer terminals but also handheld devices. These systems are able to decrease redundant data entries by communicating with other data sources, such as hospital billing systems (for demographic data), patient-completed questionnaires, internet educational and reference resources such as e-pocrates, data from other providers, and order entry for such items as added β-adrenergic receptor blocking agents, smoking cessation aids, statins, aspirin, and even walking prescriptions. Such processes allow more than one provider to record information for the evaluation and to effectively communicate that information to the anesthesiologist or other care-givers at different times or locations.
The provider systems are still evolving. Ease of use and ease of data entry are very important for acceptance by physicians, who seem to have a very short tolerance time for clumsy system performance. Also, because typing on a computer can interfere with the physician-patient interaction, a conscious effort must be made to move away from the computer and address the patient directly, so that he or she does not "feel like a number."
Laboratory systems, pharmacies, and administrative billing systems have led the way in providing on-line data systems. These data may be integrated into the collection process if suitable computer interfaces can be constructed. More problematic is the incorporation of older records, as well as notes or other nondigital data from outside sources.
Figure 25-17a
Sample checklist for determining which preoperative laboratory
tests should be obtained. CPK, creatine phosphokinase; ECG, electrocardiogram; HCT,
hematocrit; Hgb, hemoglobin; PT, prothrombin time; PTT, partial thromboplastin time;
SIADH, syndrome of inappropriate antidiuretic hormone secretion; SMA 6 and SMA 12,
simultaneous multichannel analyses of 6 and 12 blood components, respectively; WBCs,
white blood cells.
Databases consist of an "engine" and an interface. The engine is the computer program that actually stores and retrieves information. Many commercial engines such as Dbase, Sybase, and Access, as well as proprietary programs written by vendors, are available. Although a full discussion of the different engines is beyond the scope of this chapter, several questions that must be addressed when evaluating any database system are discussed here.
First, these systems range from intensely complex, requiring a full-time programmer available for maintenance and modifications, to relatively user-friendly, perhaps maintainable by a physician who has some background in database management. When selecting a system, the potential buyer must consider the resources required for maintenance.
Second, these systems vary in their ability to exchange information with other database systems, such as a hospital-based information system. I believe that ease of exchange, both now and in the future, should be considered in the selection of a system. Third, the hardware that implements the system also warrants some consideration, although to a lesser degree than the previous two criteria. Today, processor speed and storage on desktop systems are making the selection of hardware a less critical issue than before.
Finally, thoughtful design of the interface system is very important, because this feature determines the everyday ease and usefulness of the system. Many installed medical information systems (estimated at 30% in 2004, down from 45% in 1999) have not been successful, even though they are technological sound. A large portion of these failures is due to nonacceptance by users.
Clinicians want to be able to easily specify how they get information in and out of the system (the interface), and to be able to modify that interface without being charged for each change in format of a report. Programs such as Access include so-called "drag-and-drop" formats that allow the user to easily create new input and report forms. Other systems may require programming assistance or, in the case of a proprietary system, a more
The hardware systems that run these programs have become more affordable and accessible, and many of these systems can run on personal computers. However, the clinic environment is much more demanding than the office environment. The physician working in an office can tolerate a certain amount of instability and downtime on a computer, but these conditions cannot be allowed in a clinic without considerable consequences. For example, if a paper record is unavailable for one or two cases, the cases may be delayed for 20 minutes. If the server or network is unavailable, 15 ORs will be delayed for 20 minutes each.
Information systems in a clinic environment should have adequate backup in order to meet the clinical demands of being available 24 hours a day. The cost of having adequate backup and supporting personnel must be included in any planned installation. Maintenance costs for information systems can be as high as 70% of a system's total cost over the its life span. Security concerns and documenting conformance with the HIPAA make the use of a personal computer use more difficult, even as distributed databases become more useable and more desirable (given equal security systems, it is more time-consuming and more difficult to hack into 4,000 computers than one central database).
If an electronic information system only collects information and duplicates the current paper system, the cost-benefit advantage is likely to be marginal. Furthermore, it is highly unlikely that merely automating the collection of data will produce significant time savings for those involved. The benefits become clear when data are analyzed and disseminated.
The first benefit concerns the reporting of data to facilitate evaluation, discussion, and subsequent care of the patient. By necessity, the paper record had to report data in the order it was recorded. Reexamination of the paper report may show that simple electronic reformatting of the displayed information, allowing key points to be emphasized, is beneficial. If the reports are available on-line, they no longer need a linear presentation. Only the
Figure 25-18a
Sample patient questionnaire for determining which preoperative
laboratory tests should be obtained.
The computer can provide reference material on unfamiliar conditions or drugs—sometimes even automatically, as in the case of warnings about possible drug interactions. Also, when an institution wants to guide and track use of its resources, the computer can suggest preferred clinical pathways. The system may actually do an initial analysis of the data, based on algorithms, to assist in clinical decision-making. The ASA and other standard-setting organizations, including the committee advising the state of Maine, recognize the importance of clinical judgment and the interaction between patient and physician ( Fig. 25-19 ).
Such systems have the advantages of always posing a comprehensive set of questions to the patient and of not forcing the clinician to compromise comprehensiveness because of time constraints. Beers and colleagues[277] showed that such a system did a better job of obtaining accurate and complete patient histories than time-pressed clinicians. Computer-assisted analysis of data will expand as methods are developed for computers to use all the data they have accumulated.
Research can be another opportunity for benefit.[278] [279] [280] However, as an initial warning, it should be emphasized that these systems build large data sets rapidly, and the temptation to search the data for a correlation is both strong and misguided. Correlations found in this manner are often misleading and, only if supported by logic, may at best provide the basis for asking more formal questions. Pursuit of these questions, in a prospective fashion, can be rewarding. Fleisher and colleagues[281] used such a system to determine that the risk for similar Medicare patients undergoing similar operations was substantially
Figure 25-19
Two standards published by the American Society of Anesthesiologists
(ASA) in the ASA Directory of Members 1998. (A) Statement on routine preoperative
laboratory and diagnostic screening. The routine use of laboratory or diagnostic
screening tests is not an essential part of the preanesthetic evaluation of patients.
(B) Basic standards for preanesthesia care. The history, physical examination,
and chart review are essential when obtaining tests as part of preanesthesia care.
(From American Society of Anesthesiologists The ASA Directory of Members,
1998. Park Ridge, IL. American Society of Anesthesiologists, 1998.)
Through improved coding and billing, the use of preoperative databases has been shown to improve economies and cost recovery.[26] [27] [280] These improvements in cost-recovery mechanisms, along with improvements in laboratory utilization and reimbursement for medically indicated ICD-9-coded test ordering, may justify the cost of implementing a system in the context of a preoperative clinic (see later).
One of the main reasons for instituting an electronic record is to ensure availability of the collected and analyzed data on the day of surgery. At one institution, 20% of manually completed forms were being lost in the process. This loss
Institutions vary in the frequency with which patients return, but having the basic information for a patient from a previous procedure can be time-saving for subsequent elective operations. It can even be life-saving in the event of an urgent procedure for which old records are unavailable and the patient cannot provide a history.
Every discussion of information systems must consider security and confidentiality. The same access that is necessary for efficient dissemination of information can have negative effects on patient trust and a significant legal, financial, and criminal impact if improperly used. It is crucial that the system has both controlled access and accurate logging of use to minimize the opportunity for inappropriate use and to identify abuse. Users of the system must be informed of the significant and strictly applied penalties for improper use of the records. This area requires that expert assistance, both technical and legal, should be obtained by anyone who is implementing an information system containing patient data.
|
|
|
|
|
|
|
|
|
|
|
|
|
