team - organization

1ffe70d02a0b3b3e70b80fafd5f89760428b.pdf

Home >English homework help >Article writing homework help >team - organization

INNOVATIONS IN PRIMARY CARE

Can Health Care Teams Improve Primary Care Practice? Kevin Grumbach, MD Thomas Bodenheimer, MD

THE NOTION OF A HEALTH CAREteam is as rarely challenged inprinciple as it is achieved in prac- tice. Currently, a resurgence of interest in team-based care is evident. The In- stitute of Medicine has called for a New Health System for the 21st Century with primary care teams playing a central role.1 The quantum leap in the com- plexity of tasks prevents physicians alone from coping with the scope of practice. The imperative of cost containment leads provider organizations to favor lower- paid clinicians over physicians. The de- mand for quality encourages primary care to add caregivers with skills that physicians may not possess.

Will the primary care team come of age in the 21st century? This article in the series “Innovations in Primary Care” critically examines the primary care team. We begin by addressing the fun- damental question: “What is a team?” We describe 2 practice organizations that highlight key characteristics of high- functioning teams. Discussions of teams generally include a consideration of 2 major issues: “Who is on the team?” and “How does the team function?” This ar- ticle primarily focuses on the second question.

Groups and Teams In health care settings, individuals from different disciplines come together to care for patients: the surgeon, nurse, and anesthesiologist in the operating room; the oncologist, radiation thera- pist, and surgeon for patients with can- cer; and the physician, medical assis- tant, and receptionist in the primary

care office. These groupings conform to one definition of a patient care team as “a group of diverse clinicians who communicate with each other regu- larly about the care of a defined group of patients and participate in that care.”2

But is a group of people who happen to be thrown together in a surgical suite or primary care office truly a team?

Dr R works in a private practice that includes herself and one other general in- ternist. She begins her 20-minute visit with Mr H by thumbing through the chart to find the dates and results from his most recent hemoglobin A1c, low-density lipo- protein cholesterol, eye examination, and prostate-specific antigen tests. The of- fice has a medical records clerk never trained to perform these tasks. Dr R then spends 5 minutes comparing the medica- tion bottles brought by Mr H with her

chronic medication list. Reviewing the health maintenance form, she leaves the room to request a medical assistant to draw up pneumonia and influenza im- munizations, finding the medical assis- tant sitting at her desk waiting for in- structions about what to do next. Returning to the examination room, Dr R learns that Mr H has been unable to ob- tain an appointment with the urologist for a prostate biopsy; she promises to ar- range the appointment herself. As Mr H leaves, Dr R realizes that she did not need

Author Affiliations: Department of Family and Com- munity Medicine, University of California at San Fran- cisco. Corresponding Author: Thomas Bodenheimer, MD, Bldg 80-83, San Francisco General Hospital, 1001 Potrero Ave, San Francisco, CA 94110 (tbodie @earthlink.net). Section Editor: Drummond Rennie, MD, Deputy Editor, JAMA.

In health care settings, individuals from different disciplines come together to care for patients. Although these groups of health care personnel are gen- erally called teams, they need to earn true team status by demonstrating team- work. Developing health care teams requires attention to 2 central ques- tions: who is on the team and how do team members work together? This article chiefly focuses on the second question. Cohesive health care teams have 5 key characteristics: clear goals with measurable outcomes, clinical and administrative systems, division of labor, training of all team members, and effective communication. Two organizations are described that demon- strate these components: a private primary care practice in Bangor, Me, and Kaiser Permanente’s Georgia region primary care sites. Research on patient care teams suggests that teams with greater cohesiveness are associated with better clinical outcome measures and higher patient satisfaction. In addi- tion, medical settings in which physicians and nonphysician professionals work together as teams can demonstrate improved patient outcomes. A num- ber of barriers to team formation exist, chiefly related to the challenges of human relationships and personalities. Taking small steps toward team de- velopment may improve the work environment in primary care practices. JAMA. 2004;291:1246-1251 www.jama.com

a medical degree to accomplish any of the tasks performed during the medical visit.

Few people would argue that Dr R, the medical records clerk, and the medi- cal assistant—although working to- gether to care for the same patients— truly function as a team.

A consideration of teams evokes 2 questions: “Who should be the play- ers on the team?” and “How can the players act as a team rather than as a collection of individuals?” A familiar team is a football squad. Football teams need the right mix of players. A 22- member team with 11 quarterbacks and 11 defensive linebackers would win few games. Similarly, a primary care prac- tice with 3 physicians and no recep- tionist, medical assistant, or billing clerk is seldom a winning combination. Even with the right mix of players, a foot- ball team with no plays, no practice ses- sions, and no game plan would be un- likely to land in the Super Bowl. In fact, such a “team” is not truly a team but simply a group of individuals. Even though groups of health care person- nel thrown together in an office, clinic, or hospital floor are generally called teams, they need to earn true team sta- tus by demonstrating teamwork.3

A simple definition of team may help to distinguish unstructured groups vs organized teams: “A team is a group with a specific task or tasks, the accom- plishment of which requires the inter- dependent and collaborative efforts of its members.”4 The football example clarifies the difference between a group and a team:

It is naive to bring together a highly di- verse group of people and expect that, by calling them a team, they will in fact be- have as a team. It is ironic indeed to realize that a football team spends 40 hours a week practicing teamwork for the two hours on Sunday afternoon when their teamwork really counts. Teams in organizations sel- dom spend two hours per year practicing when their ability to function as a team counts 40 hours per week.4

A Brief History of Primary Care Teams in the United States The general practitioner of the early 20th century was a lone ranger. Black

bag in hand, he treated and comforted patients, often in their homes. As of- fice practice emerged, the first pri- mary care team was husband and wife, the wife serving as receptionist, bill- ing clerk, and bookkeeper. As practice became more complex, nonphysician tasks became further subdivided into receptionist, medical assistant, and bill- ing clerk, a pattern found in the myriad of small practices dotting the United States.

In 1915, teams of physicians, health educators, and social workers were cre- ated at Massachusetts General Hospi- tal’s outpatient department. Primary care team models were developed at New York’s Montefiore Hospital in 1948 and Yale in 1951.4 The Neighborhood Health Center program of the 1960s devel- oped primary care teams in some early health centers.5,6 Larger group prac- tices also incorporated a diverse comple- ment of health professionals into teams.

Despite these efforts, primary care teams did not become the dominant paradigm. One obstacle was the “over- whelming barriers of disciplinary ter- ritoriality and systems inertia.”7 Pay- ment systems failed to reimburse work performed by nonphysician profes- sional members of the team, undermin- ing the financial viability of teams. Per- haps most critical was the failure to articulate the objectives of a team model and to provide evidence on the advan- tages of team models for achieving these objectives. The “team meeting”— lengthy sessions in which each team member offered his or her perspective on a patient and family—became em- blematic of these failures. A physician veteran of 1970s team practice writes,

Because our goals were so lofty, we needed to spend enormous amounts of time together to explore all the intricacies and nuances of comprehensive health care. . . . The central organizing event was the pre-clinic meeting, usually held at one of a number of local restaurants. Quite regu- larly, critical decisions were made at this time. The most memorable of these was res- taurant selection for the next week . . . the patients/families? No one really knows how it was for them (R. Goldschmidt, MD, writ- ten communication).

At the start of the 21st century, do well-functioning primary care teams ex- ist in the United States? Two case stud- ies highlight organizations striving to- ward high-performing teams. One setting is a small private office, the other a large group practice.

Two Contemporary Primary Care Teams Dr Charles Burger. Charles Burger is a private practitioner in Bangor, Me. From a distance, this remarkable pri- mary care practice resembles thou- sands of physician offices throughout the country. Upon entering the office door, it is clear that—within a tradi- tional practice setting—Dr Burger has created a smoothly functioning pri- mary care team. The entire office func- tions as one team—2 physicians and 2 nurse practitioners are the clinicians, complemented by medical assistants, greeters, receptionists, and schedul- ers. The practice is financially stable and is busy, with each clinician seeing 23 to 30 patients per day. The following case typifies how the team model works:

Ms P called Dr Burger’s office com- plaining of recurrent abdominal discom- fort after eating. The receptionist con- sulted her computerized triage protocol and told Ms P to come the same day. When she arrived, the greeter, already aware of the patient’s problem, gave her a medical his- tory questionnaire specifically related to abdominal pain, which Ms P filled out in the waiting room. Ms P met with the medi- cal assistant who checked her vital signs and quickly entered her questionnaire re- sponses into the computer. Ms P then saw the physician, who reviewed the history, performed a relevant physical examina- tion, and consulted a diagnostic software program. Discussing the options with Ms P, the physician and patient decided on a diagnostic and treatment plan. Ms P then met with the scheduler, who arranged labo- ratory and ultrasound studies.

Dr Burger’s staff members were all trained at a 15-week course in quality management at a nearby college. Greet- ers, receptionists, and schedulers (who are cross-trained) also received 6 weeks of in-office training.

INNOVATIONS IN PRIMARY CARE

All clinical processes in Dr Burger’s office are guided by a system. The prac- tice has adopted advanced access sched- uling, offering patients same-day ap- pointments. For years, the office has tracked demand and can predict how each day will unfold. On Mondays, heavy with telephone calls, more staff act as receptionists and few scheduled appointments are made.

Whereas in most offices, reception- ists are not trained to properly triage pa- tients into emergency, urgent, and rou- tine categories, Dr Burger designed a triage system that receptionists consult on every telephone call. When Ms P called with abdominal complaints, the receptionist pulled up the gastrointes- tinal screen on the triage protocol, which prompted a series of questions includ- ing pain severity and presence of vom- iting, diarrhea, black and/or bloody stools, or fever. In the case of positive answers, the protocol tells the recep- tionist to send Ms P to the emergency department. For milder symptoms, an appointment is made, perhaps with pre- visit laboratory studies. The interac- tion is routed to Ms P’s medical record and a clinician’s e-mail in-box.

Most communication is routinized by the office’s clinical systems. Team mem- bers do not attend endless meetings. In- coming calls are routed to the e-mail in- box of the appropriate team member. Urgent messages are delivered in per- son. Diagnostic studies go to the appro- priate e-mail in-box and the medical rec- ord. The well-trained medical assistants order clinical preventive studies based on the patient’s age and sex. Clinic goals and performance measures are commu- nicated to all staff by posters promi- nently displayed in the office.

Kaiser Permanente in Georgia At the opposite end of the primary care spectrum is Kaiser-Permanente’s large delivery system. In 1997, Kaiser- Permanente’s Georgia region (KP/ Georgia) developed primary care teams with several goals: increased patient sat- isfaction, improved Health Employer Data and Information Set scores, and lowered costs.

This group practice model cur- rently consists of 9 primary care of- fices with 25 teams. Each team has 3 to 5 clinicians (physicians, nurse prac- titioners or physician assistants), 2 reg- istered nurses, 1 to 2 receptionists or clerks, and 6 to 7 licensed practical nurses or medical assistants, provid- ing care to a panel of 8000 to 15 000 pa- tients. Prior to the rollout of the team structure, clinicians and staff received training in team-oriented care.

Patients view their clinician, not the team, as their primary caregiver, but are aware that a nonphysician clinician may provide care for acute problems or if the physician is not available. Eighty-five percent of visits are handled by a cli- nician on the patient’s team.

Kaiser-Permanente’s Georgia team, like Charles Burger’s practice, has well- defined systems and protocols for all clinical processes, including triaging telephone calls, reviewing and inform- ing patients of laboratory and x-ray re- sults, making referrals, and renewing prescriptions. One registered nurse is the advice nurse, answering patient questions and triaging patients who telephone or drop in. The other regis- tered nurse is the team coleader, work- ing with the physician coleader to solve day-to-day problems, ensure that clini- cal systems are functioning well, and supervise team members.

Each team receives a budget based on the number of patients on the team’s panel with risk adjustment according to age and disease severity. Initially given limited decision-making au- tonomy, teams demonstrating effec- tive self-management are allowed flex- ibility in staffing mix and division of labor. Teams can decide if they want more physicians, more nonphysician clinicians, or more support staff in their personnel mix. Some teams delegate chronic care management functions to licensed practical nurses and medical assistants; others are less successful in this redesign. Each team decides how chronic disease registries are used to im- prove its panel’s outcome measures. Some use the registries extensively, oth- ers minimally.

Each team receives a quarterly re- port on team functioning, patient sat- isfaction, staff satisfaction, and clini- cal quality measures, enabling KP/ Georgia’s central leadership to assess each team’s functioning and allowing each team to compare itself with other teams.

Building Teams What are the features that distinguish the teams of Charles Burger and KP/ Georgia from the dysfunctional work- ing group of the fictional Dr R? The conceptual work of several scholars has highlighted 5 key elements of team building: clear goals with measurable outcomes, clinical and administrative systems, division of labor, training, and communication (BOX).3,7-9 Both Dr Burger’s practice and the KP/Georgia teams exemplify these 5 elements. Both of these practices have concrete goals and measure their performance in reaching these goals, eg, patient satis- faction, good clinical outcomes, and in the case of KP/Georgia, cost reduc- tion. Both of these institutions have es- tablished detailed systems to accom- plish the tasks that all primary care practices must fulfill. They have con- structed a division of labor so that each team member knows, and is well- trained to accomplish, the role he or she must play in performing each task. Dr Burger’s practice illustrates a creative approach to division of labor by devis- ing nontraditional positions, such as the “greeter,” and by delegating to recep- tionists and medical assistants some tasks that are typically performed by cli- nicians. Practice systems, division of la- bor, and training are missing elements in the fictional practice of Dr R. While Dr Burger’s practice makes a substan- tial investment in staff training, Dr R—like many primary care prac- tices—puts new employees to work after a scant 2-hour orientation. In the practices of both Dr Burger and KP/Georgia, communication is accom- plished via systems and protocols and by face-to-face, minute-to-minute conversations rather than by lengthy meetings.

INNOVATIONS IN PRIMARY CARE

Research on Health Care Teams Is there evidence that building a cohe- sive primary care team is worth the ef- fort—for better quality of care, lower costs for equivalent quality, or im- proved workplace satisfaction?

To the extent that clinical care in- volves groups of people working to- gether, research indicates that a group with better teamwork tends to per- form better than one lacking team- work. Although studies differ in how they measure teamwork, most ap- proaches attempt to capture the types of characteristics listed in the Box. In a variety of industries, research has found that team cohesiveness is asso- ciated with effectiveness in carrying out the team’s tasks, even though there is not always an association with im- proved productivity.3,10-14 A team- oriented culture in intensive care units is associated with better technical qual- ity of care, lower length of stay, and improved relationships with family members but not with improved risk- adjusted mortality.15

Two recent studies of general prac- tices in England demonstrated that bet- ter teamwork and team climate are as- sociated with better processes of care for patients with diabetes16 and better continuity of care, access to care, and patient satisfaction.17 Primary care teams in Spain with clear goals and healthy communication did better than less cohesive groups on the outcomes of patient-perceived quality and pa- tient satisfaction.18 One exploratory study found that better relationships with practice staff were predictive of greater job and career satisfaction among physicians.19 Researchers have studied the KP/Georgia teams de- scribed above. Initial findings suggest that teams with higher “collaborative clinical culture” scores have superior patient outcomes, including better pa- tient satisfaction and better control of diabetes and hyperlipidemia.20

In addition to research on team- work, other studies have investigated the effect of team composition on prac- tice outcomes: Who should be the play- ers on the primary care team? One mo-

tivation for adding team members is to conserve expensive physician labor through substitution of other person- nel for physician effort. Baldwin7 of- fers the “general dictum that services (tasks) should be performed at the low- est level of professional training, which

leaves those with greater training or re- sponsibility free to perform tasks or to solve problems for which they are uniquely equipped.”

Most formal research on substitution in primary care has examined the role of nurse practitioners and physician assis-

Box. Key Elements of Team Building

1. Defined Goals Overall organizational mission statement

Examples: Improvement of patient’s health Reduction in barriers to access to care Improvement in practice’s financial performance Physician and staff satisfaction

Specific, measurable operational objectives Examples:

At least 80 of diabetic patients in practice will have hemoglobin A1c lower than 8

Ninety percent of people calling for a nonurgent appointment will receive the appointment within 1 week

Practice will achieve a targeted level of practice revenue Each team member will achieve an explicitly identified goal for personal

professional development

2. Systems Clinical systems

Examples: Procedures for providing prescription refills Procedures for informing patients of laboratory results

Administrative systems Examples:

Procedures for making patient appointments Policies on how decisions are made in the medical practice

3. Division of labor Definition of tasks Assignment of roles (Determining which people on the team perform which

tasks within the clinical and administrative systems of the medical practice

4. Training Training for the functions that each team member routinely performs Cross-training to substitute for other roles in cases of absences, vacations,

or periodic heavy demands on one part of the team

5. Communication Communication structures

Examples: Routine communication through paper and electronic information flow Minute-to-minute communication through brief verbal interactions among

team members Team meetings

Communication processes Examples:

Giving feedback Conflict resolution

INNOVATIONS IN PRIMARY CARE

tants. Two recent meta-analyses pro- vide evidence that nurse practitioners can deliver care of equivalent quality to that delivered by primary care physi- cians,21,22 with the caveat that most stud- ies reviewed included small numbers of clinicians and few examined long-term outcomes for patients with chronic ill- ness or complex conditions. In con- trast, there is a lack of rigorous research on other types of substitution in pri- mary care, comparing how different staff- ing mixes of registered nurses, licensed practical nurses, and medical assistants affect patient outcomes. Nor has re- search assessed clinical outcomes for the substitution occurring in Dr Burger’s practice through delegation of greater clinical responsibilities to medical assis- tants and receptionists.

Research is also inconclusive about whether substitution of personnel with lower salaries in primary care always translates into lower cost per visit. Al- though several studies indicate that the use of nonphysician clinicians can reduce costs in primary care prac- tices,23-27 some of these studies have been criticized for not fully accounting for nonphysician clinicians seeing fewer pa- tients per hour and working fewer hours per week than primary care physi- cians.2 8 A recent study of the KP/ Georgia teams that carefully accounted for visit productivity and work effort con- cluded that teams that made greater use of nurse practitioners and physician as- sistants relative to physicians had lower overall team labor costs per visit.29

Configuring team personnel is not just a matter of substitution for eco- nomic benefit. Another objective is en- hancement of clinical performance. Team members may contribute unique talents that enhance the skill mix of the practice. Wagner2 has argued that nurses’ training makes them better than physicians at following chronic care management protocols. Nurse practi- tioners may also have better patient education and communication skills than do physicians. Numerous stud- ies suggest that multidisciplinary clini- cal teams produce clinical outcomes su- perior to those achieved by “usual care”

arrangements, with many of these stud- ies evaluating the addition of nurses, so- cial workers, psychologists, and clini- cal pharmacists to teams.2,30-40

A limitation of these studies is that few examine team models for the varied problems that predominate in primary care practice. Many of these studies focus on hospitalized patients or in- vestigate teams designed for a well- demarcated population or condition, such as frail elders or patients with dia- betes. Confounding variables are plen- tiful, such as the quality of personnel on one team vs another and the resources available to one team vs another.41

The Problems With Teams If teams are such a good idea, why aren’t they more prevalent? Teams have some inherent drawbacks related to their added organizational complexity. As team size increases, the transaction costs of interpersonal communication in- crease exponentially and may over- take the benefits of teamwork.42 Team size may have a U-shaped relation to teamwork; too few or too many team members reduce effectiveness.8 One study suggests that 6 team members is the optimal size; teams with greater than 12 members are too large.43 Teams also require dealing with the challenges of human relationships and personali- ties.44 While some team members may shine as initiators, clarifiers, or encour- agers, others may play negative roles as dominators, blockers, evaders, and rec- ognition seekers.3

Despite evidence that teams may en- hance clinical performance, teams may conflict with other important practice values. Delegating tasks to other team members may erode work satisfaction for the generalist physician attracted by the idea of personally delivering compre- hensive care. Greater team size may in- terfere with patient preferences for con- tinuity of care with a single clinician.42

The undifferentiated and varied na- ture of clinical problems in primary care makes team building especially chal- lenging. A single specialty “service line” practice will find it relatively easy to de- lineate tasks and define roles, com-

pared with a primary care practice fac- ing a more diffuse array of clinical tasks.

Finally, financial incentives matter. Economic disincentives are promi- nent under current fee-for-service pay- ment policies; an office visit with a phy- sician or nurse practitioner, but not with a medical assistant, is billable, ne- gating the economic benefit of the prac- tice of substitution.

Introducing Teams Into Primary Care How can primary care physicians take the first steps toward creating more ef- fective teamwork? Building a cohesive primary care team begins with an as- sessment of one’s own working group, using the Box as a guide. Does the prac- tice have clearly articulated clinical, business, and work environment goals with measurable outcomes to assess im- provement? One English general prac- tice successfully pioneered a process of engaging clinicians and staff to agree on practice goals.45 Once the goals are for- mulated, does the practice have the best mix of personnel to meet the goals?

Do detailed systems exist to routin- ize practice tasks, for example, how pa- tient telephone calls are triaged, how laboratory and x-ray results (normal or abnormal) are communicated to pa- tients, and how refills for different cat- egories of prescriptions are handled? Does each team member have clearly defined tasks within these systems and is each well-trained to perform those tasks? In the vignette about Dr R, the medical records clerk could have been trained to maintain a flow sheet with patients’ laboratory data and the medi- cal assistant could have been trained how to make specialty appointments for elderly patients unable to navigate the health care system. Training team mem- bers takes place on the job and does not require additional funding. The train- ing does require physicians to spend time up front, an investment that should save physicians like Dr R substantial time over the long run.

Could nonphysician personnel sub- stitute for physicians in performing some tasks, thereby decompressing

INNOVATIONS IN PRIMARY CARE

physician work load? Practices fre- quently underuse the capabilities of re- ceptionists and medical assistants. In the case of Dr R, a medical assistant could have been trained in comparing the patient’s medication list with the pills the patient was actually taking, sav- ing Dr R 5 minutes of the medical visit. Low-cost investment in staff training— either on-the-job or in local commu- nity colleges—can unleash the full po- tential of team members.

Barriers to team development are

considerable. A predator of the pri- mary care team is the hamster.46 “Ham- ster health care”—the rapidly revolv- ing treadmill upon which so many clinicians find themselves—creates a state of mental exhaustion that frus- trates attempts at planning and coop- eration. Though a well-functioning team with a clear division of labor might relieve physicians of some of their workload, finding the time to partici- pate in team development is difficult for physicians. Whether or not a primary

care practice chooses to focus on team development as a major innovation, many practices may benefit by intro- ducing or improving one or more com- ponents of high-performing teams— clear goals with measurable outcomes, defined tasks and roles, clinical and ad- ministrative systems with a clear divi- sion of labor, and effective communi- cation. Making time to step off the treadmill to invest in team planning may yield long-term benefits in the form of an improved work environment.

REFERENCES

1. Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Wash- ington DC: National Academy Press; 2001. 2. Wagner EH. The role of patient care teams in chronic disease management. BMJ. 2000;320:569-572. 3. Fried BJ, Rundal TG, Topping S. Groups and teams in health services organizations. In: Shortell SM, Kaluzny AD, eds. Health Care Management. Albany, NY: Del- mar Thomson Learning; 2000. 4. Wise H, Beckhard R, Rubin I, Kyte AL. Making Health Teams Work. Cambridge, Mass: Ballinger Pub- lishing Co; 1974. 5. Lashof JD. The health care team in the Mile Square area, Chicago. Bull NY Acad Med. 1968;44:1363- 1369. 6. Wise H. The primary care team. Arch Intern Med. 1972;130:438-444. 7. Baldwin DC. The Role of Interdisciplinary Educa- tion and Teamwork in Primary Care and Health Care Reform. Washington, DC: Health Resources and Ser- vices Administration, Dept of Health and Human Ser- vices; 1994. 8. Cohen SG, Bailey DE. What makes teams work: group effectiveness research from the shop floor to the executive suite. J Manage. 1997;23:239-290. 9. Rubin IM, Beckhard R. Factors influencing the ef- fectiveness of health teams. Milbank Mem Fund Quart. 1972;50:317-335. 10. Bettenhousen KL. Five years of group research: what we have learned and what needs to be ad- dressed. J Manage. 1991;17:345-381. 11. Gully SM, Devine DJ, Whitney DJ. A meta- analysis of cohesion and performance. Small Group Res. 1995;26:497-520. 12. Cannon-Bowers JA, Oser R, Flanagan DL. Work teams in industry: a selected review and proposed framework. In: Swezey RW, Salas E, eds. Teams: Their Training and Performance. Norwood, NJ: Ablex Pub- lishing; 1992. 13. Vinokur-Kaplan D. Enhancing the effectiveness of interdisciplinary mental health treatment teams. Adm Policy Ment Health. 1995;22:521-530. 14. Seashore S. Group Cohesiveness in the Indus- trial Work Group. Ann Arbor: University of Michigan Institute for Social Research; 1954. 15. Shortell SM, Zimmerman JE, Rousseau DM, et al. The performance of intensive care units: does good management make a difference? Med Care. 1994; 32:508-525. 16. Stevenson K, Baker R, Farooqi A, Sorrie R, Khunti K. Features of primary health care teams associated with successful quality improvement of diabetes care. Fam Pract. 2001;18:21-26.

17. Campbell SM, Hann M, Hacker J, et al. Identify- ing predictors of high quality care in English general practice: observational study. BMJ. 2001;323:1-6. 18. Goni S. An analysis of the effectiveness of Spanish primary care teams. Health Policy. 1999;48:107-117. 19. Williams ES, Konrad TR, Linzer M, et al. Refining the measurement of physician job satisfaction. Med Care. 1999;37:1140-1154. 20. Roblin DW, Kaplan SH, Greenfield S, Roberts MH, Jacobs LD, Carlton DG. Collaborative clinical culture and primary care outcomes. In: Program and ab- stracts of the annual meeting of the Academy for Health Services Research and Quality; June 23-25, 2002; Washington, DC. 21. Brown SA, Grimes DE. A meta-analysis of nurse practitioners and nurse midwives in primary care. Nurse Res. 1995;44:332-339. 22. Horrocks S, Anderson E, Salisbury C. Systematic review of whether nurse practitioners working in pri- mary care can provide equivalent care to doctors. BMJ. 2002;324:819-823. 23. Flood AB, Fremont AM, Jin K, Bott DM, Ding J, Parker RC. How do HMOs achieve savings? the ef- fectiveness of one organization’s strategies. Health Serv Res. 1998;33:79-99. 24. Scheffler RM, Waitzman NJ, Hillman JM. The pro- ductivity of physician assistants and nurse practition- ers and health work force policy in the era of man- aged healthcare. J Allied Health. 1996;25:207-217. 25. Grzybicki DM, Sullivan PJ, Oppy JM, Bethke AM, Raab S. The economic benefit for family/general prac- tices employing physician assistants. Am J Manag Care. 2002;8:613-620. 26. Hooker RS. A cost analysis of physician assis- tants in primary care. JAAPA. 2002;15:39-42, 45, 48. 27. Venning P, Durie A, Roland M, Roberts C, Leese B. Randomized controlled trial comparing cost effec- tiveness of general practitioners and nurse practition- ers in primary care. BMJ. 2000;320:1048-1053. 28. DeAngelis CD. Nurse practitioner redux. JAMA. 1994;271:868-871. 29. Roblin DW, Howard DH, Becker ER, Adams EK, Roberts MH. Use of midlevel practitioners to achieve labor cost savings in the primary care practice of a MCO. Health Serv Res. In press. 30. DeBusk RF, Miller NH, Superko Hr, et al. A case- management system for coronary risk factor modifi- cation after acute myocardial infarction. Ann Intern Med. 1994;120:721-729. 31. Aubert RE, Herman WH, Waters J, et al. A ran- domized controlled trial of nurse case management within an HMO to improve glycemic control in patients with diabetes. Ann Intern Med. 1998;129:605-621.

32. Wagner EH, Glasgow RE, Davis C, et al. Quality improvement in chronic illness care: a collaborative ap- proach. Jt Comm J Qual Improv. 2001;27:63-80. 33. Sommers LS, Marton KI, Barbaccia JC, Randolph J. Physician, nurse, and social worker collaboration in primary care for chronically ill seniors. Arch Intern Med. 2000;160:1825-1833. 34. American College of Physicians-American Soci- ety of Internal Medicine. Pharmacist scope of prac- tice. Ann Intern Med. 2002;136:79-85. 35. Hanlon JT, Weinberger M, Samsa GP, et al. A ran- domized, controlled trial of a clinical pharmacist in- tervention to improve inappropriate prescribing in el- derly outpatients with polypharmacy. Am J Med. 1996; 100:428-437. 36. Bogden PE, Abbott RD, Williamson P, Onopa JK, Koontz LM. Comparing standard care with a physician and pharmacist team approach for uncon- trolled hypertension. J Gen Intern Med. 1998;13:740- 745. 37. Gattis WA, Hasselblad V, Whellan DJ, O’Connor CM. Reduction in heart failure events by the addition of a clinical pharmacist to the heart failure team. Arch Intern Med. 1999;159:1939-1945. 38. Leape LL,Cullen DJ, Clapp MD, et al. Pharmacist participation on physician rounds and adverse drug events in the intensive care unit. JAMA. 1999;282: 267-270. 39. Halstead LS. Team care in chronic illness: a criti- cal review of the literature of the past 25 years. Arch Phys Med Rehabil. 1976;57:507-511. 40. Wells KB, Sherbourne C, Schoenbaum M, et al. Impact of disseminating quality improvement pro- grams for depression in managed primary care: a randomized controlled trial. JAMA. 2000;283:212- 220. 41. Schmitt MH, Farrell MP, Heinemann GD. Con- ceptual and methodologic problems in studying the effects of interdisciplinary geriatric teams. Gerontolo- gist. 1988;28:753-764. 42. Barr DA. The effects of organizational structure on primary care outcomes under managed care. Ann Intern Med. 1995;122:353-359. 43. Starfield B. Primary Care. New York, NY: Oxford University Press; 1998. 44. Lencioni P. The Five Dysfunctions of a Team. San Francisco, Calif: Jossey-Bass; 2002. 45. Adelaide Medical Centre Primary Health Care Team. A primary health care team manifesto. Brit J Gen Pract. 1991;41:31-33. 46. Morrison I, Smith R. Hamster health care: time to stop running faster and redesign health care. BMJ. 2000;321:1541-1542.

INNOVATIONS IN PRIMARY CARE