Managed care and health systems have investedheavily in improving care for such chronic diseasesas diabetes mellitus and coronary arterydisease. Chronic medical conditions are common, morbid,and expensive. Increased adherence to guidelinesoften improves patient outcomes and decreases costs ofcaring for these conditions. By their very nature, chronicconditions are ideal to address from a qualityimprovement or behavior change standpoint. Generally,the long time interval over which the diseases progressgives health systems months or years to "do the rightthing" for individual patients. Actions take place at multiplepoints along the continuum of care–before, during,or after medical visits–and are carried out byvarious members of the healthcare team, any one ofwhom can be the focus of quality improvement efforts.
Care for acute conditions, on the other hand, is moredifficult to affect. Acute illnesses tend to be brief andinvolve limited interactions between the patient and thehealthcare system. Any successful attempt to changeacute care must affect clinician and patient behavior atthe time of the healthcare visit.
Despite this difficulty, acute conditions remainattractive targets for care improvement efforts, mainlybecause of their frequency and cost. Acute respiratoryinfections (ARIs), for example, are the most commonreason patients seek care in the United States, representingabout 7% of the 890 million ambulatory visits in2002.1 ARIs are also the number one reason physiciansprescribe antibiotics in the United States. ARIs accountedfor 47% to 56% of all adult and 75% to 80% of all pediatricantibiotic prescriptions between 1991 and 1999.2Much antibiotic prescribing for ARIs is inappropriateand exposes patients to potential adverse drug events,increases bacterial resistance, and increases cost. Thetotal cost, direct and indirect, of ARIs is at least $40 billionannually in the United States.3 If health systems areto improve care for acute conditions, then ARIs are anatural place to start.
Four articles in this issue of the describetechniques used to affect care processes for ARIs.Brunton et al developed a clinical guideline for the treatmentof patients with acute exacerbations of chronicbronchitis.4 Hutt et al sought to improve the care ofpatients with nursing home-acquired pneumoniathrough organizational changes and provider education.5 Ahmed et al aimed to increase influenza vaccinationrates among high-risk patients through the use ofpostcard reminders to patients and an influenza "toolkit" for employers.6 Finally, Greene et al sought toreduce inappropriate antibiotic prescribing for acutesinusitis by using physician education, physician profiling,and financial incentives.7 As in the wider universeof studies on improving care processes, the results hereare mixed and the magnitude of benefit was modest.
Other previously successful interventions toimprove the quality of care for ARIs–specifically toreduce inappropriate antibiotic prescribing–haveincluded academic detailing, formulary restrictions, andmultidimensional interventions involving patients andclinicians. As previous studies have shown, and thepresent articles indicate, multifaceted interventionsseem more likely to change behavior. However, many ofthese interventions have been expensive and time-limited;it is unclear how sustainable any improvementswill be.
The Qualities of an Effective Health InformationTechnology Solution
Another solution, health information technology, hasthe potential to improve care for ARIs in a cost-effective,sustainable way if done right. For example, use ofa template in an electronic medical record can captureclinical information in a standardized way, automaticallyimport patient information (such as medication andproblem lists), and potentially decrease inappropriateantibiotic prescribing by providing real-time clinicaldecision support. Such a template could also facilitatethe printing of relevant patient handouts and provideeasy access to relevant literature. At PartnersHealthCare, of which Brigham and Women's Hospital isa part, we are developing an ARI "Smart Form" for ourelectronic medical record that incorporates many ofthese features. However, significant barriers remain.
The most critical barrier is also the most obvious: forthe majority of acute problems, decision support mustbe used during an office visit in order to be effective. Inaddition, clinicians will only accept real-time decisionsupport if it is fast, simple, intuitive to use, anticipatesclinicians' needs, and is integrated into clinicians'workflow.8 Clinical decision support systems cannottake the clinician more time and should, in fact, bedesigned to save the clinician time. For example, cliniciansshould not have to interact with decision support,separately make an order, and then document thatorder. These three steps can and should be collapsedinto one.
Furthermore, clinical decision support systemsshould not impair patient-physician communication orsap physician autonomy. Easy ways to override decisionsupport software must be provided; if the softwareis too restrictive, it will be rejected. When clinical decisionsupport identifies an error of commission (such asprescribing an inappropriate antibiotic), it needs tooffer a useful alternative. Clinical decision support systemsshould be able to accommodate information fromvarious practice members (eg, nurse with vital signs,physician with history and physical examination data)and accommodate the workflow of the entire office staffwithout turning clinicians into secretaries. Such systemsshould also provide something for the patient(handouts with detailed recommendations).
To encourage the use of clinical decision supportsystems, health information technology needs to providethe clinician with self-evident, immediate benefits.Such self-evident benefits could include automaticallyabstracting and organizing patient data; combiningupdating of the medical record, ordering, and documentationinto one step; facilitating writing of customizedpatient letters and patient instructions; andidentifying the preferred medication in a class for apatient's insurance coverage. In addition, providinginformation that a certain level of evaluation and managementdocumentation has been met could result inincreased reimbursement for clinicians, thus encouraginguse. Finally, clinicians need adequate training tounderstand how to use any new information technologies,especially ones that impact their workflow.
An Even Larger Role for HealthInformation Technology
In the future, fully functioning, electronic medicalrecord-based ARI-care templates have the potential toaffect care on a broader scale by more seamlessly integratingthe public health and the personal healthcaresystems. Such templates can provide real-time surveillancedata to public health officials related to naturallyoccurring disease clusters, novel respiratory pathogens,or bioterrorist attack. Information can also flow in theother direction, from the public health system to theclinician, so that information from disease surveillanceprojects could be fed in real time to clinicians, to letthem know if there really is "something going around."Up-to-date information from a department of healthabout the local prevalence of influenza could be displayedwithin a patient's chart, and integrated with apatient's signs and symptoms, to show the probabilityof that patient having influenza.
Much of this may sound pie-in-the-sky at present.Indeed, all this is predicated on the use of sophisticatedelectronic medical records, and the adoption ofhealth information technology in the ambulatory settinghas been disappointingly slow. Resistance usuallycomes because health information technology is perceivedas being too large an investment or as makingthe office or clinic more difficult for already busy clinicians.However, the potential gains to managed careorganizations, health systems, clinicians, and patientsare enormous in terms of improving safety, quality,and cost.
We have made significant strides in improving thequality of care for chronic conditions, and, of course,more needs to be done. However, while we continueto improve the quality of care for chronic diseases,we must nevertheless move forward to introduce systemsof care that improve quality for acute conditionsas well.
AcknowledgmentI thank Jeffrey Schnipper, MD, MPH for his critical review of thiseditorial.
From the Division of General Medicine, Brigham and Women's Hospital and HarvardMedical School.
Dr. Linder is supported by a Career Development Award (1 K08 HS014563) from theAgency for Healthcare Research and Quality.
Address correspondence to: Jeffrey A. Linder, MD, MPH, Division of GeneralMedicine, Brigham and Women's Hospital, 1620 Tremont Street, BC-3-2X, Boston, MA02120. E-mail: firstname.lastname@example.org.
1. Woodwell DA, Cherry DK. National Ambulatory Medical Care Survey : 2002Summary. Advance data from vital and health statistics; no 346. Hyattsville,Maryland: National Center for Health Statistics. 2004.
Ann Intern Med.
2. Steinman MA, Gonzales R, Linder JA, Landefeld CS. Changing use of antibioticsin community-based outpatient practice. 2003;138:525-533.
3. Fendrick AM, Monto AS, Nightengale B, Sarnes M. The economic burden ofnon-influenza-related viral respiratory tract infection in the United States. 2003;163:487-494.
Am J Manag Care.
4. Brunton S, Carmichael BP, Colgan R, et al. Acute exacerbation of chronicbronchitis: a primary care consensus guideline. 2004;10:689-696.
Am JManag Care.
5. Hutt E, Reznickova N, Morgenstern N, Frederickson E, Kramer AM. Improvingcare for nursing home-acquired pneumonia in a managed care environment. 2004;10:681-686.
Am J Manag Care.
6. Ahmed F, Friedman C, Franks A, et al. Effect of the frequency of delivery ofreminders and an influenza tool kit on increasing influenza vaccination ratesamong adults with high-risk conditions. 2004;10:698-702.
Am J Manag Care.
7. Greene RA, Beckman H, Chamberlain J, et al. Increasing adherence to a community-based guideline for acute sinusitis through education, physician profiling,and financial incentives. 2004;10:670-678.
J Am Med Inf Assoc.
8. Bates DW, Kuperman GJ, Wang S, et al. Ten commandments for effectiveclinical decision support: making the practice of evidence-based medicine a reality. 2003;10:523-530.