-collaboration and the quality of health care delivery

8/10/2019 -Collaboration and the Quality of Health Care Delivery

1/30

6/13/11: CH added refs to Hoff in 2 places and added that ref to biblio

Collaboration and the Quality of Health Care Delivery*

Charles Heckscher

Rutgers University

School of Management and LaborRelations

[email protected]

609-644-4689

Saul Rubinstein

Rutgers University

School of Management and LaborRelations

Niclas Erhardt

Maine Business School

Boniface Michael

California State University, Sacramento

Draft: May 14, 2011

*We would like to acknowledge the invaluable help of our partners in this research: Joel Cantorand Mary Ellen Cook, of The Rutgers Center for State Health Policy; Sebastian Palmeri, MD, ofthe UMDNJ-RWJMedicalSchool; and John McCarthy, of the Rutgers SMLR PhD program.
mailto:[email protected]:[email protected]


2/30

Collaboration and the Quality of Health Care Delivery

AbstractHospitals face major obstacles to reform of their organizations despite a widely-acknowledgedand urgent need for improved coordination. This paper explores the change process in fourhospital cardiac care units through in-depth case studies. We find that change efforts are shapedand constrained by three complex cultural patterns: traditional-professional, administrative,and collaborative. Attempts to impose administrative culture generally create conflict, whiledevelopment of collaborative culture is slow and difficult but shows some promise.


3/30

3

Introduction

The healthcare delivery system has increasingly struggled to coordinate complex relationsamong highly specialized and professionally autonomous actors. Many studies have identifiedinterprofessional relations as central to the achievement of safety and quality of care. These

studies suggest that complex care is best achieved through collaboration and engagement ofmultiple professional stakeholders, through mechanisms such as multidisciplinary rounds orinter-professional teams (Asch et al., 2005a; DAmour, Ferrada-Videla, San Martin Rodriguez, &Beaulieu, 2005b; Gittell, 2008; Gittell et al., 2000c; San Martn-Rodrguez, Beaulieu, DAmour, &

Ferrada-Videla, 2005d).

Yet it has also become apparent that such reform is extremely difficult. Although a few positivecases of collaborative organization like the Mayo Clinic and Intermountain Health Care havebeen widely trumpeted (Berry, 2004; Maccoby, 2006), the vast majority of the US health caresystem remains fragmented and inefficient (Davis, Schoen, & Stremikis, 2010).

Why is it so hard, in this domain of great societal importance and attention, to implementcollaborative reforms which have strong face validity as well support from as formal researchfindings? We look for the answer in patterns of culture patterns of work relations and mutualexpectations that are difficult to change using a three-part typology of organizational culturethat organizes and helps to explain the dynamics of the process. In this view, moves towardsgreater cooperation involve not just behavioral changes but also systemic shifts in interactionpatterns; the path is not straight but requires extremely complex changes in attitudes andinteractional recalibrations.

We explores these dynamics through a detailed study of work systems around the treatment ofcongestive heart failure in four New Jersey hospitals. This set of cases enables us to see in somedetail, and in very different settings, how work relations evolve and affect the provision of careon multiple dimensions: quality, efficiency, patient satisfaction, and long-term impact.

We first outline three basic types of professional culture as applied to the organization ofhospital care. Then we describe our research approach and describe the four cases in detail,with the theoretical categories as an organizing tool. Finally, we draw conclusions from thesecases about the dynamics of change, resistances, and outcomes of different approaches.

Three professional cultures and the delivery of health care

Scholars have used widely varying conceptualizations of organizational culture; Scheins(1990)review finds there is presently little agreement on what the concept does and shouldmean.We will go back to the sociological classics for our definition, using the term to refer tosystems of social norms and values (Parsons, 1963, 2007). Norms are the expectations thatgovern relations among differentiated roles and statuses; they define practices, the way things

get done around here, (Deal & Kennedy, 2000). Values by contrast refer to ideals shared byeveryone. These two dimensions differentiated relational patterns and shared values overtime become tightly integrated systems that are difficult to penetrate or change. Conversely,new systems, without such integration, are generally unstable. Used in this way, culture isequivalent to the term community, which is also defined by shared norms and values (Adler,Kwon and Heckscher, 2008b).


4/30

4

A number of scholars have observed that the professional medical culture is often an obstacle tochange in health care, particularly because of its strong norm of autonomy. Hoff (1999)documented the cat and mouse game of management change initiatives and physicianresistance. Leape and Berwick (2005) blamedthe slow rate of adoption of safety improvementson a culture of medicine centered on a tenacious commitment to individual, professional

autonomy. Ramanujam and Rousseau (2006) see described health care organizations as theperfect storm of organizing difficulties render[ing] leadership weak and vulnerable todemands of multiple professions seeking to assert control over their own professional practice.Moreover, Scottie, DAmour, & Moreault(2002) show have shown how intractable theseproblems can be: after 25 years of effort to develop an interdisciplinary culture in Quebeccommunity health care centers, they found only modest changes in beliefs and values andcontinued tension with a professional or disciplinary logic.

We will argue, however, that the culture referred to in these studies is not the universalmedical culture but a particular, albeit common, type. Adler et al. (2008) suggest, congruentwith a number of others, that the dominant mode is closely related to traditional craft culture,with doctors as the master craftsmen of the system (Kieser, 1989; Krause, 1999). This type is

marked by strong and stable status hierarchy with physicians at the top, as recipients of specialknowledge and socialization, and operating with essential autonomy modified only by modestpeer self-governance (Freidson, 2001; Goode, 1957; Light & Levine, 1988). Nurses and otherroles in the system have their own special functions nurses are generally seen as caring, forexample, in a role-definition that goes back at least to the medieval period but they arestrongly subordinate to doctors (Daiski, 2004; Weinberg, 2003).1These key elements of theculture cause major resistance to interprofessional collaboration (Gugan, 1995; Hall, 2005; Light& Levine, 1988; Royal College of Physicians of London, 2005).

Adler et al identify two further forms of community. The first centers on a commitment toimpersonal rules and bureaucratic hierarchy. They call thisgesellschaft,following Toennies; wewill call it for clarity adminstrative community. This pattern is based on classic Weberian

rationality and formalized authority; it has been seen in healthcare over the last few decades inthe rise of large hospitals with strong administrations, often led by non-medical professionals,who seek to impose standardized standards and procedures (Engel, 1970; Scott, Ruef, Mendel,& Caronna, 2000; Weinberg, 2003).

The final model is the collaborative one, still in an early stage, which draws professionalstogether as knowledge expert working cooperatively to focus on the patient. It has been calledby various names, including network, collaborative,and team-based, andseen asfocused on the creation of interaction and dialogue among actors with differing capabilities(Heckscher, 2007; Powell, 1990; Sabel, 1991). It is institutionalized through deliberate processesof dialogue and interactive planning, and it requires a shift in values from an emphasis onindividual autonomy and performance towards interdependence and contribution to a sharedmission.

Table 1 sketches the main characteristics of these three types of organizational culture and theirhistorical development in business and medicine.

Thus our core research interests are to understand the dynamics of the creation of collaborativecultures: how efforts to create them interact with traditional expectations, the conflicts and


5/30

5

resistances that result, and how these may be overcome. Along the way we will also test claimsthat collaborative culture improves the quality of care.

----------------------------------------

Insert Table 1 About Here

-----------------------------------------

Method

We focused on congestive heart failure because it is a complex syndrome involving a delicatebalance of diet, exercise, fluid retention, medication, and physical therapy both in the hospitaland in outpatient care. This usually involves multiple hospital stays and many providers

making complex judgments, from surgeons through general practitioners, pharmacists, nurses,respiratory therapists, and transport staff. It is also one of the major causes of mortality and oneof the greatest costs in the US health system.2

In order to explore cultural and relational patterns we use a multimodal, primarily qualitativeapproach that captures many dimensions of organization, including interactions and reactionsto change by different groups in cardiac units.

Selection of Hospitals

We chose research sites from all New Jersey hospitals based on two selection criteria:performance and payer mix. We excluded hospitals that had specialized transplant centers.Aside from this simple structural issue, we knew nothing about the organizationalcharacteristics of our sample hospitals before we began our research.

Performance was important as an outcome measure because we hypothesized that certaincultural types would yield better performance than others. We also wanted to control for payermix because we were concerned that the many unmeasurable differences between poorhospitals, whose patients have difficulty in paying and who are generally highly costconstrained, and rich ones with wealthier patients and higher budgets, might overwhelm thealmost everything else including the organizational factors we are studying.

Comparable performance data was scarce and difficult to obtain. First, we gathered in-hospitalmortality data from the Centers for Medicare and Medicaid Services and adjusted it for riskusing APR-DRG 3M software, Version 15. Second, we used measures of process consistency

gathered by the State of New Jersey reflecting percentage of completion of four key processes in2005. Because neither measure is an adequate measure of performance by itself, we soughthospitals who scored high or low on both. Of the 76 hospitals for which we obtained completedata on both dimensions, 11 hospitals scored in the top quarter on both dimensions, and 8scored in the bottom quarter on both. For payer mix we used data on the percentage ofMedicare and self-pay patients, using the top and bottom 25% on this dimension for our pool.


6/30

6

From this group we were able to gain access to one hospital for each cell in our desired matrix.Thus we had one hospital with wealthy patients and poor results, and one with the opposite poor patients and good results as well as two that followed the more expected pattern.

--------------------------------------


--------------------------------------

Data Collection

Data were collected over a 12 month period, by a research team including two professors oforganization theory, a nurse, a physician, and two advanced doctoral students trained inqualitative method techniques. We used the following research methods:

Semi-structured interviews: We interviewed a total of 85 informants across the four hospitals.

These were chosen to select a cross-section of the functions that affect the care of heart failurepatients physicians, nurses, pharmacists, housekeeping, telemetrists, case managers, labtechnicians, and others. See table 3 for more detail.

----------------------------------


----------------------------------

Observations: In all cases we conducted observations about the flow of work in the main heartfailure unit. In all but one we observed medical rounds (the fourth one did not hold predictablerounds). In one case we shadowed a doctor in the emergency room during an entire shift.

Analysis of Data

We adopted a coding procedure, as suggested by Miles and Huberman (1994) and Glaser(1978), that involves cycles of individual coding, team refinement of concepts, and elaborationof new codes. The burden for the case-based researchers is to avoid cherry picking, orhighlighting data that serve to support preconceived notion of what we might discover. Thus,attention to disconfirming data is critical. In order to strengthen the validity of our data andresults, we used the constant comparison method (Trochim, 1989), where our workingtheoretical framework of relational systems were checked and rechecked against new evidenceto detect potential disconfirming evidence. To further improve construct validity and reliabilitywe adopted a triangulation strategy (Mathison, 1988), relying on multiple informants,

observations, and archival data.

The cases

We will begin with brief analytic descriptions of the four cases, and then take up in more detailthe lessons we can draw about the sources of the difficulties of change and some potentiallypromising avenues. Leaders in two of these cases were trying to impose administrative controlsand measures, and two were seeking to build more collaborative relations. The latter were moresuccessful, but faced many difficulties.


7/30

7

Case 1 Riverside The persistence of traditional relations

At Riverside the traditional medical culture remained largely intact, with long-term andstable relationships, despite gentle efforts by administrators to introduce more rationalizedadministrative approaches. This was a medium-sized community hospital with a relativelyhigh-end patient population (few self-pay or Medicaid,) and therefore not under severe

pressure for change. Nevertheless, like all hospitals, it was experiencing continual pressures forimproved efficiency and cost control. It was in the lowest quarter of the states hospitals on ourmeasures of quality of cardiac care.

This case represents an inwardly focused traditional craft model, with good cohesion withinunits derived from long employment tenure but weak multi-disciplinary communication andcollaboration. It presented an interesting paradox. The care providers both physicians andnurses were generally quite happy with the standard of care and with the workingenvironment, embracing the sense of family solidarity; but the objective measures of processand outcome were very poor.

All actors both doctors and nurses, as well as other support staff held strongly to traditional

ways of doing things, despite clear problems. Physicians were very resistant to properdocumentation or using the computer systems; they wanted to be able to enter information intheir own particular ways and without pausing to make sure their writing was legible. Thissometimes led to errors of interpretation by other providers. Nurses and doctors who hadworked together for a long time had worked out some of these problems, though not all; butthose who were newer or more peripheral were unable to communicate effectively. A nursereflected:

The people who have been here for a while, they pretty much know who to call andyou dont get so many breakdowns. We get much better communication. It's mostlywhen new people come in that we have the breakdowns.

There were stories which vividly illustrated the costs of these limits. For example, a

phlebotomist a relatively low-level medical specialist focused on drawing blood told us of arecurring problem he faced:

In certain patients for whom it was difficult to find a good vein, a permanent arterialline was implanted; but the line might be hidden by a sheet or the patients position,and there was no information about it on the printout that was given to me. Wewould sometimes find a very weak patient, unable to respond, and spend manyminutes trying to find a clear vein, although an arterial line was available.

The interviewer (slightly horrified) asked the phlebotomist at this point whether he had talkedwith physicians or administrators. His response was: I havent been in the right opportunity toraise that yet.

Riversides administrators had tried a number of efforts to improve communication andcooperation, but with little success. Problem-solving committees, for example, were unpopular:the common refrain was that there was too much paperwork and too many meetings. Anurse reflected:

[Problem solving] is done informally: the nurses and the pharmacists talk togetherand we try to see how we can improve it. But then we take it to a committee, and thenthey'll put their input into it and it's on to an ad hoc committee, and then all of a


8/30

8

sudden it gets lost. And then we're told no, we don't have the money for that rightnow, and you shouldn't have asked me. And then you're back to square one, trying tofigure out another way to do it

Interdisciplinary rounds had been instituted in an effort to improve coordination, but thesewere very ineffective. Information presented at the rounds was incomplete and out of date;

nurses did not have accurate knowledge of the patients under discussion. Rather than servingas forums for interdisciplinary problem-solving, these rounds centered on case managers andconsultants who sought to identify people who had exceeded their length of stay and should bedischarged, physicians were peripheral.

Case managers, who had been hired to try to coordinate care through a patients entire stay,were defensive and complained of being disliked by everyone; The doctors, said one, feelwere challenging their medical decision-making.

Similarly, administrative attempts to introduce cost values into the mission statement werelargely ineffectual. Most people said that they had read the mission but didnt really think aboutit much. One nurse was more blunt: My mission is caring for the patient, and get out of my

way.In general, there was continued resistance to administrative reforms. Care providerscomplained that the administration was trying to impose alien standards and did not respectthem. We used to be a family, said a nurse; now were two families.

We both heard about and directly observed regular subversion, through informal resistance, ofexternal pressures for accountability. Doctors, for example, were able to avoid administeringfeedback forms to patients who they feared would not be positive, and physicians failed tocomply with external audit recommendations for practice improvements.

There was little awareness of the financial situation, even among nurse managers. Nurses andclinical staff viewed patient care as primary and cost pressures am external imposition to be

battled: one said, I dont know much about the finances, but I do know we have to fight toothand nail for every cent.

The informal sense of unity was limited by strong professional boundaries. In the absence oflong-stranding personal relations, physicians did not expect their medical orders to bequestioned by the nurses, pharmacists, or other clinical staff. Similar professional defensivenessmarked communication between nurses and different clinical staff including pharmacists,psychologists and rehabilitation. And even in the best circumstances there was considerablereluctance to violate the traditionally-defined status hierarchy.

Thus multiple weaknesses of the traditional craft pattern of professional culture were evident,especially:

Limitations on open dialogue about problems: such discussions were limited to informalnetworks and blocked for new people or those, like phlebotomists and case managers,who had not made their way into the family circle

Difficulty in balancing care with other priorities, especially cost. Everyone felt veryfocused on providing the best care, and felt good about this; they largely rejectedcompeting concerns.


9/30

9

Inconsistency of procedures and difficulty in introducing new approaches andtechniques.

Ineffectiveness of focused problem-solving processes, preference for informal and adhoc communication.

Case 2Lowell: The failure of administrative reform

At Lowell a new administration, pursuing widely-recommended efforts at greater processstandardization and accountability, aggressively confronted traditional medical communities,precipitating severe conflict that at times broke into open rebellion, with intense frustration onboth sides.

This was a medium-sized (about 300 beds), short term community hospital with a good localreputation as a quality hospital, despite a low-income patient mix. However, in recent years, ithad gone through several ownership changes and its performance had declined on a range ofquality and efficiency measures. In the year before our research it fell into the lower quarter ofresults on both our mortality measure and the state-published process data. New administrators

had felt as of two years before that there needed to be significant improvement to avoid a futurecrisis. Thus they had initiated a series of change efforts, an attempt to standardize physicianprotocols on a few key practices with strong evidence of their effectiveness; and seeking toreduce length of stay by hiring active case management consultants. Unlike at Riverside, theLowell administration pushed hard to overcome resistance with a combination of educationand stronger accountability and incentives.

The result was a fine example of open conflict between the administrative andcraft/professional orientations. Physicians were highly resistant to the new systems. Forexample, the administration had introduced interdisciplinary rounds led by case managers, butthe physicians generally did not attend. Both nurses and doctors generally expressed hostility tothe effort; in their view, as one nurse put it: It's not about helping people, it's about making

money!An administrator expressed equally intense frustration: Physicians are a major blockfor any change in this hospital.

Doctors were not, however, the only focus of difficulty: nurses, and their union, had engaged inopen conflict with the administration around cost-cutting moves and reassignments. Theadministrators insisted that cost reduction was essential and that the nurses had refused tonegotiate around these issues. Following a work stoppage, many nurses had left and thehospital had taken to using large numbers of agency nurses, hired on a temporary basis. Theconstant change in agency staff, and tension with regular nurses, exacerbated problems ofcoordination. Regular nurses saw agency nurses as having poor work ethic, low motivation andlack of dedication to patients; agency nurses complained vociferously that the regular staff wereunprofessional and sloppy in their procedures.

To reduce length of stay, the administration had hired case managers whose entire focus was onmanaging this issue. These case managers led the interdisciplinary rounds and steered thediscussion for each patient to the question of whether they met the criteria for insurancereimbursement and, if not, how to move them out. The physicians were peripheral to rounds, ifthey attended them at all. Despite this effort, length of stay had changed little in the previousyear. Both nurses and doctors generally expressed hostility to the effort.


10/30

10

The administration had hoped to develop some control over physician procedures by hiringhospitalists doctors employed directly by, and accountable to, the hospital. But nurses, as wellas the hospitalists themselves, viewed the communication between doctors and hospitalists asineffective, with the community doctors protecting their own relation to patients, so that all thathad been accomplished was the addition of an extra element that further confused efforts at

coordination.Patient advocates had run into similar obstacles: the head of this function felt cut out ofdecision-making and frustrated by a lack of respect. She had been unable to achieve thecoordination she sought:

Often times patient feel that there is a disconnect: consulting told them something,another doctor comes in and say you can leave today, but then the hospitalist comes inand says, I don't know, I'm waiting for a testI think they get the impression thatsometimes the right hand does not know what the left hand is doing.

Lack of coordination was a continuous theme as expressed by nurses:

The hospitalist sees the patient in the emergency room, and it's consult, consult,

consult. Then the specialist comes in, sometimes reading those initial orders,sometimes not reading those initial orders, and everything gets changed.

Another nurse added:

Every member of the team really should be at some part during the day comingtogether to discuss whatever the goals and the plan for the patient so you don't missanything. And I think that would definitely help the length of stay because everyonewould know what's going on. We have different parts of the team meeting at differenttimes.

In sum, the strong intervention of the administration to try to get control of a highlyunsatisfactory situation had undermined the web of traditional relationships between

physicians and nurses, but without replacing it with either bureaucratic accountability orreliable mechanisms of collaboration. The physicians attitude of professional autonomy hadremained intact, leading to open mutual hostility between the administration and thephysicians. Nurses, caught in the middle, felt they had lost their traditional responsibility forpatient care. Lack of coordination was a constant theme of the interviews. All objectivemeasures, internal and external, showed very poor results in congestive heart failure treatment.

Case 3 Hightown: Limited collaboration in a wealthy hospital

At Hightown, a strong physician leader largely succeeded in transitioning from a traditionalto a collaborative culture, with markedly improved quality results, though with manyremaining tensions at the boundaries.

This was a large community hospital with a relatively wealthy payer population and little costpressure. It was among the very best hospitals in the state on both mortality and processmeasures for heart failure.

Six years before it had initiated a major program for cardiac care, with some governmentsupport, bringing in a strong doctor cardiologist who had formed a heart failure team with tolead it and two Advanced Practice Nurses (APNs) to support him. Key elements in thisinitiative included:


11/30

11

Frequent and highly organized interdisciplinary rounds, including the physician leader,the APNs, a social worker, physical therapist, staff nurses, dietician and of the supportstaff.

Formalized multidisciplinary problem-solving committees and processes, reviewing thequality of care and making suggestions for improvement and standardization. These

were done within the framework of a Strategic Mission statement very similar to thatat Riverside.

The empowerment of nurses as key coordinators of care. They were able to access fullinformation about the patients and were given support and training by the APNs so thatthey could have informed discussions with the community doctors: as Stephanie, anurse put it, When you know what you're talking about, they do respond to that.

We observed in detail ways in which these collaborative mechanisms improved the quality ofcare. The interdisciplinary rounds were particularly effective. Although these rounds wereformally the same as those at Lowell and Riverside, their actual operation was very different.During our observation in during the round, nurses from the floor came to present their set of

patients; an Advanced Practice Nurse was sitting at a computer with all the patient data in frontof her; other specialists, including the cardiologist who led the heart failure team, sat around theroom. Very frequently problems of coordination were uncovered and fixed, and in many casesefforts were made to diagnose the source of the problem and prevent a reoccurrence. Twoexamples from among many:

A nurse, said that her patient needed a change in medication; the physician said he hadalready done that. The nurse replied: No, it hasnt changed. After some discussion andchecking the records they found that the Emergency Room had failed to transmit theorder. They immediately fixed the issue for that patient, identified the ER physicianresponsible at that time, and assigned someone to speak to him.

Another nurse suggested that a patient could be sent home, although he was still proneto falling. Loraine, the social worker intervened: That wont work, his family is inupstate New York, he needs more support. A decision was made to keep him longerdespite the length-of-stay guidelines.

The cardiologist-APN team provided constant leadership for care and change, but in a way thatwas generally embraced by the nursing community rather than rejected like the hospitalists andcase managers at Lowell.

The collaborative efforts had been extended beyond the hospital setting. Outpatient care wastightly integrated with the hospital staff. The two APNs had also established ongoing relationsto homecare providers and offered advice and support to home care nurses without any directreimbursement. This clearly had a major impact. Home care, according to the nurses, was

ordinarily very poorly coordinated: nurses had very little information on the treatment history,doctors came very infrequently, and nurses were not expected to initiate calls to the doctorsexcept under very clear circumstances, and never to suggest changes in treatment. We heardabout the development of new processes to improve the information sheets passed from thehospital to the home care providers.

There were many formal problem-solving committees, and though there were some mixedreviews, there was far less of a negative feeling than at Riverside. They were able at times to go


12/30

12

well beyond the kind of informal workarounds that were common at Riverside. One nursedescribed a committee to look at the handling of inventory:

I emailed people; they joined the team because it was a very big problem, everybodyhad an issue with it and they wanted to come to the table because they had somethingto say. And once they got together, they heard the other people and they started seeing

the bigger picture it wasn't just one thing.

We also heard of incidents of open dialogue in which lower level providers successfullypushed back against physicians, not through informal personal relationships but through aneffective use of collaborative mechanisms. A home care nurse recounted,

There was a woman who was discharged [but] not dried out as well as she could be. Ithought, If she doesn't start taking off the water she will end up back in the hospital. Itold the primary physician; He said Look, we've done all the diuretics we can on thisperson. And I looked at the chart and I said, Well, they didn't try Viroxillin. Butthen I said, Before I start suggesting things I don't know enough about she only hasone kidney, I don't want to upset her kidney function I should talk to the expert toget their opinion.

This nurse spoke to one of the hospital APNs, who confirmed her view and encouragedher to go back to the doctor; the doctor agreed to try this avenue.

In sum, this was a promising case of collaborative relations, bringing together differingspecialties for effective patient-centered problem-solving. It was clear that the strong leadershipof the lead physician was a key element in this relative success; the role on Advanced PracticeNurses as communication links and central network players was also essential.

Nevertheless, there remained important obstacles that continued to threaten the effort, and tolimit the scope of its impact on practices in other parts of the hospital. One of the most visiblewas continued resistance by physicians with visiting privileges, who dealt with most of the

patients in the cardiac unit: they continued to protect their own autonomy and ways of doingthings and were resistant to suggestions emerging from the multidisciplinary rounds. The leadcardiac doctor kept himself in the background in part to minimize the resistance of his fellowdoctors, and the discussions in the rounds often concerned how to bring around communitydoctors who were not supporting the effort.

The nursing hierarchy was another major locus of tension: the lead physician had someimpolitic things to say about the head nurses unwillingness to adjust to the needs of the unitand her insistence on consistency of rules across the entire hospital, which made innovationvery difficult. More generally, other parts of the hospital had not rushed to embrace thesuccessful innovations of the cardiac unit: the higher-level administrators whom weinterviewed echoed the findings of other studies cited above that the norm of autonomy

continued to overshadow the need for information-sharing and learning. We will come back tothese and other impediments to change in the discussion section.

Case 4 United: Limited collaboration in a cost-constrained hospital

United was a particularly interesting case because it had high quality outcomes, includingsignificant improvement in recent years, despite a very low-income patient population and aconstrained budget. It was pursuing a collaborative effort rather similar to that at Hightown,though at an earlier phase of development and with some modifications required by cost


13/30

13

considerations. Despite the difficult patient mix, the hospital fell into the top 25% on ourmeasure of adjusted in-hospital mortality and on the states assessment of process consistencyfor 2006 and 2007.

Because it was a teaching hospital, the organization, unlike the other cases in our study, rarelydealt with community (outside) physicians. Physicians were employees of the hospital and

therefore more easily available when needed for consultation.

For three years before our research the hospital had undertaken a major initiative to improvecardiac care. Like Hightown, it hired a new cardiac physician leader of the cardiology unit, whointegrated the Fellows into a strong and tight-knit leadership team similar to the heart failureteam at Hightown. Two physician Fellows played a role somewhat similar to the AdvancedPractice Nurses at Hightown, helping to educate and spread the principles of the change effortconsistently through the unit.

This case represents the collaborative type in an early stage of development, centering heavilyon the strong leadership of a the new head of the department, a cardiologist who emphasized astrong focus on patient care. One of the physicians noted:

If there is any one [key to the quality of heart failure care] it is him being the trueleader in building a team. He is the champion who holds people accountable, in a niceway. He has helped improve nurse-doctor relations. He brings professionals fromdifferent disciplines into it, too. We are doing well because of this level of teamwork.

Like Hightown, this unit relied on a set team of highly skilled practitioners who helped spreadand reinforce the message; unlike Hightown, these key agents in this case were Fellows(physicians in undergoing specialized training in cardiology) rather than Advanced PracticeNurses.3

The effort focused at first on improving process measures based on National Quality Forumsstandards for treatment of heart attack and congestive heart failure. These scores had improved

significantly over the previous two years. This focus on process measures, according to the headof cardiology, was a springboard for ensuring compliance to core measures that need to bedone. By including it in the core measures it gets done. Documentation, which had beeninconsistent, became far more standardized and the main procedures could be easily checked.

Despite the strong emphasis on the lead physicians role, the effort operated on a philosophy ofinvolvement. A cross-disciplinary problem-solving committee apparently the first of its kind,at least in this part of the hospital initially studied current procedures, interviewed people,and recommended improvements. Nurses and other providers at all levels were involved in thediscussions, and nurses were empowered to carry out improvements with far less bureaucraticchecking than usual. In our interviews at lower levels we found that many, though not all, ofthe care staff bought in to the effort, were aware of the main issues involved, and supported

the leadership team.

To emphasize integrated care and coordination, two further multi-disciplinary committees hadbeen established focused on cardiac services and performance improvement. They weregenerally seen as effective. A case manager, for example, told of a team she worked on thatcollected data on major reasons for delays in discharge, identifying the key holdups astransportation, tests, and lab reports. These findings were reported to the cardiologyimprovement team and corrective action was taken. Such seemingly simple changes are


14/30

14

indicative of how difficult it is to coordinate the usual hospital procedures from different areasaround the overall care process, and how effective even a simple cross-functional team can be.

There was a consistent effort to educate staff at all levels about medical issues. Ben, the head ofcardiology said, When you are talking about quality of care you are really talking about reallytraining people how to treat heart failureand an academic program does not necessarily

mean youre going to have expertise. He emphasized an active clinical research program, drugand device trials in part because it elevates the level of expertise of every person involved,including the nurse at bedside.

For the most part dynamics in the department were characterized by an cooperativerelationship between the doctors and nurses, an emphasis on education and research in patientcare and an active approach to problem solving and conflict resolution.

There were some who did not share the general positive view. A few nurses who had beenthere a long time complained about continuing tensions between nurses and other roles pharmacists, administrators (there are too many layers)as well between older and youngernurses. Several senior nurses complained about understaffing and failure to maintain a

sufficient nurse-patient ratio: Even if we have our contract and it is specific what the ratio isbetween nurse and patient, in certain situations it still wont happen.They preferredadvancing complaints through their own nursing hierarchy and were rather dismissive of thenew committees and processes.

The majority, however, gave a very different impression more than any of our other sites, itwas like engaging two different worlds or mindsets. The majority of those we interviewed wereenthusiastic about the new leadership, engaged in the changes, and hopeful about the future.They were generally critical of the ingrown culture and convoluted work systems, butnoted that the changes brought by the new lead physician were recent, and progress was beingmade, albeit unevenly.

The four doctors we interviewed consistently expressed respect for the nurses skills. Nursesand other professionals involved in patient management generally felt that the doctors wereaccessible and receptive. Most nurses felt empowered to grab a doctor on patientmanagement issues.A dietician felt empowered to clarify and modify doctors initial dietorders. Nurses felt that they would readily communicate with the Fellow first in case ofproblems.

Budget awareness remained relatively low. One of the Lab supervisors typified the lack ofclarity about reimbursement norms for tests: I think the insurance company pays for the tests.But we dont have anything to do with that at all. We do whatever tests are asked of us. Thedepartment had nevertheless achieved reasonable cost control through focusing on criteria foradmission, patient care while being admitted and final discharge of the patient. There was an

attempt underway to focus on the admission and length of stay criteria prescribed by Medicareand Medicaid. To improve discharge planning, nurses were being empowered to coordinate thedischarge process with the attending physicians advice.

In broad outline this case represents a somewhat earlier version of the collaborative model seenat Hightown, with both its strengths (high quality of care, effective interdisciplinary problem-solving) and its weaknesses (tension with other parts of the hospital, reliance on the protectionof a single leader).


15/30

15

There were two aspects of the United effort that surprised us in a positive way. First, althoughthe role of the lead physician was (as predicted) very strong, he did not play the role ofguardian and protector against other parts of the hospital, as did his counterpart at Hightown.Rather than trying to seal off his unit as a special domain, he engaged in intensive networkingwith other areas and tried to draw them into his vision of cardiac care. There was some

evidence that this approach was having good effects. Relations to other units were relativelycooperative, and there was a smoother relationship with the nursing hierarchy than at our othersites.

Second, this initiative focused at the start on standardization of a very clear set of procedures the 8-scope processesrather than on a broader vision of relations. But this was driven by thedoctor, rather than administrators, and engaged a wide variety of actors at all levels in thepursuit of this objective. There were strong interdisciplinary elements, including two formalcommittees and, according to our interviews, effective working relations among specialists. TheUnited leaders framed this as just the first step in a longer-term effort to build a patient-centered focus, and felt that the quality improvements so far were just the beginning of thepossibilities.

Discussion

In the cultural categories sketched in the theoretical section, the two hospitals that had beenchosen for their poor performance on medical outcomes turned out to be ones with a sharp andvisible divides between craft and administrative cultures.

The traditional craft culture was most clearly seen at Riverside and was also still activethough weakened at Lowell. This orientation, with strong norms of professional autonomy anddeference to physicians, was clearly unable to coordinate the complexity of interdependenciesinvolved in the modern care of heart failure. There were many inconsistencies of process andmiscommunications among caregivers that contributed heavily to the poor medical results.They also failed to integrate values of cost control and efficiency into their daily operation andto balance them with medical quality.

Efforts by administrators to solve these problems through bureaucratic consistency wereineffective in these cases. Where the administrative effort was relatively low-key, at Riverside,it simply failed to penetrate the traditional relations and was seen as a kind of unnecessaryannoyance. At Lowell, where it was forced more aggressively, it had sparked hostility,disrupting the previous set of craft-like relations without building a more positive culture. Thiscase had the worst performance on medical care and had not reduced length of stay.Thesedynamics echo other studies of management-cntered reform in hospital settings (e.g., Hoff(1999)).

The two hospitals that had been chosen for excellent medical outcomes Hightown and

Unitedboth turned out to be engaging in deliberate collaborative culture change: they haddeveloped mechanisms for discussion and learning among staff from different professions,units, and levels, and there was evidence that these mechanisms were working effectively forproblem-solving. They were attempting, with partial though not total success, to build sharedvalues and expectations of open dialogue, in contrast to the traditional pattern of statusdeference. Both efforts were driven by strong cardiac physician leaders.


16/30

16

This alignment of performance and structure supports other research cited earlier showing thatcollaborative organization is effective in coordinating complex care regimens. Through ourinterviews and observations we were able to further strengthen this connection by showing thatmany of the reasonsfor the performance differences were closely related to these differences inthe way the different systems functioned and their capacity to manage the interdependence of

different providers. But we also saw the difficulties involved in changing cultural patterns.Though confounding variables could explain some of these differences we saw, none seemssufficient to undermine the evidence for the effectiveness of collaboration. One issue that couldbe significant is the fact the United was a teaching hospital, and the physicians were thereforenot independent of the hospital. While it seems likely that this integration is helpful, itssignificance as a major explanatory variable is weakened by at least three considerations. First,the cardiac unit had had poor results a few years before despite this structure and hadimproved them dramatically without changing it. Second, and conversely, Hightown wasequally successful with a collaborative effort even though most of its physicians wereindependent of the hospital. Finally, our observations and interviews brought out many reasonsfor the high quality at United that did not have to do with its teaching structure, but were more

centered on relations among different care occupations.We were concerned at the start that money the wealth of the patients and the hospital mightoverwhelm all other factors. We were interested to find that we were not only able to identifycases in both unexpected directions wealthy and unsuccessful, poor and successful but thatthe two such cases we studied did not have fundamentally different dynamics from theircounterparts. This suggests that while money is certainly important, the dynamics oforganization can have powerful independent effects: you can do well even with resourceconstraints, and vice-versa.

Change: Limitations and obstacles

All of these cases were seeking to change the medical delivery system in significant ways andall were encountering major difficulties. A particularly clear example concerns rounds. All fourhospitals had introduced or strengthened multidisciplinary rounds within the previous fewyears, to bring together various specialists and roles around patient needs. Thus on the surfacethey all appeared to be developing collaborative mechanisms and patient-centeredmanagement. But a closer look uncovered sharp differences in their actual practice.

At Riverside, dominated by the traditional craft / professional culture, interdisciplinaryrounds were very ineffective. Information presented at the rounds was incomplete andout of date; nurses did not have accurate knowledge of the patients under discussion.Rather than serving as forums for interdisciplinary problem-solving, these roundscentered on case managers and consultants who sought to identify people who had

exceeded their length of stay and should be discharged, physicians were peripheral. At Lowell, where there was an open battle between the traditional and administrative

orientations, the administration had hired case managers whose entire focus was on thisissue. These case managers led the interdisciplinary rounds and steered the discussionfor each patient to the question of whether they met the criteria for insurancereimbursement and, if not, how to move them out. The physicians were, again,peripheral to rounds, if they attended them at all. Despite this effort, length of stay hadchanged little in the previous year.


17/30

17

We have already described the relatively successful interdisciplinary rounds at Hightown. It isworth reemphasizing that these had taken a number of years to develop and were part of abroad effort at cultural change which included new problem-solving processes, informationsystems, and leadership. At United, which was in an earlier stage of a similar effort, the roundswere just beginning to function and were still facing many difficulties in bringing the various

departments together. Thus, those units characterized by the typical administrative-crafttension were able to implement multidisciplinary groups in name only; this technique wasactually operational only in units which had made a broader shift to collaborative systems.

Though the collaborative efforts at Hightown and United were largely successful andencouraging, they were very far from providing a complete model that could be transported toother locations. Just as the transition from craft organization to bureaucracy took many decades,the transition to full collaborative systems involves long-term problems that go well beyond thescope of what has been accomplished in these units. These cases represent indicators ofpotential, but it is important to emphasize also their limitations.

Some of these continuing issues were internal to the units. There was still some individualresistance, including nurses and sub-departments that did not get it. There was some ongoingsuspicion of efforts at standardization and cost control. These attitudes were more visible atUnited, which had only engaged in the collaborative course for three years, than at Hightown,where many of the internal issues had been worked out after six years.

But the more important and intractable problems by far were at the boundaries. These were in asense hothouse experiments that did not fit comfortably with many other aspects of theirenvironments. Difficulties, some of which have been mentioned above, included:

At Hightown most physicians were not employed by the hospital: it was clear that manywere not integrated with the collaborative approach and often resentedrecommendations from the interdisciplinary teams.

Both units saw the nursing hierarchy as a source of problems because of staff allocationsthat did not take account of the collaborative teams and relations.

At United there were some mentions of union rules that defined nursing roles andresponsibilities narrowly and did not support their involvement in problem-solving.

Though administrators were generally supportive, they applied considerable pressure atHightown to limit unreimbursed activities even when they clearly contributed to qualityand perhaps also to long-term cost control. One extended discussion was about two staffmembers dedicated to telephone followup with patients to make sure they werereceiving proper after care. Members of the cardiac unit were convinced that this activityhad contributed to the lowering of readmission rates, but administrators wanted to seehard data that this cost was being recovered.

In general, insurance regulations were totally external and seen as operating accordingto a foreign logic. The case manager at United and the Advanced Practice Nurses atHightown tried to play the role of protector and buffer against these demands.

Given all these obstacles, it is not surprising that the collaborative efforts required unusualleadership, which cannot be counted on for a large-scale change strategy.


18/30

18

Both of the collaborative physician leaders advocated a strategy of building a highlyautonomous cardiac unit, shielded from the rest of the hospital, in order to develop a newculture and the highest skills (though the United leader had been forced to partially abandonthis approach). This is typical of early phases in organization change efforts, and it is known tocreate the danger that the innovation will become isolated and will eventually die of neglect

even if it is successful on its own terms (Berg, Appelbaum, Bailey, & Kalleberg, 1996; Goodman,1980; Heckscher, 2007, p. 213; Rubinstein & Kochan, 2001). In the hospital setting the strategy ofautonomy creates other problems as well. The high incidence of co-morbidities heart failurepatients very often have other diseases as well means that focusing on cardiac care may notmaximize the effectiveness of total care. And this approach also increases costs because itreduces the ability to allocate beds flexibly a constant source of tension with administrators.

A further element to be worked out is the role of administration. Organization theory wouldsuggest that this needs to be a clear role in managing large-scale systems. We did not see a goodsolution in any of our cases. In the worst cases administrators were seen as interfering withtraditional relations; in the best ones they operated more quietly and supportively in thebackground. But in no instance was there a positive and shared understanding of the ways in

which professional and administrative roles could work together and how each of themcontributed to the overall picture.

Beyond this, for the collaborative approach to become widespread through the health caresystem would require large-scale changes in attitudes that do not happen easily. It would needto overcome resistance not only from traditions of professional autonomy, but also from thegrowing administrative pressure for tighter controls.

Some positive lessons for change

A collaborative route to process consistency

In the business world, companies passed through a long and highly conflictual imposition

of administrative or bureaucratic systems on the previous craft occupations around the turn ofthe twentieth century. Our two less successful cardiac units, Lowell and Riverside, echoed thathistory of conflict between the administrative and craft orientations. At Riverside professionalautonomy had (so far) blocked efforts at needed standardization and control, while at Lowellthere was open conflict.

But Hightown and United provided some evidence for another route: they achieved manybenefits of bureaucracy through collaborative means. These were the units that achieved thebest results on process consistency and the greatest improvements in length of stay. Rather thanimposing standards from the top, or even through representative committees, they managed toengage people at all levels. At Hightown, as discussed above, there was considerable unityaround balancing care with efficiency. At United, there was a clear understanding among many

we interviewed that the improvement of process consistency was important for quality care,and that they could contribute actively as professionals to that improvement.

There are analogies in other companies. Total Quality programs, for instance, sometimes workin this way (though they are more often imposed from above). Toyota has come to exemplifythis kind of highly participatory focus on process consistency what (Adler & Borys, 1996) havecalled an enabling bureaucracy. But these efforts have been built on an already-strongfoundation of bureaucratic systems and values. When the Toyota approach is extended to


19/30

19

hospitals, it tends to come into conflict with the mindset of professional autonomy (Gawande,2007).

The role of physicians as change agents

It seems significant that in both of the best cases the administrators played very much of abackground role, and the chief cardiac physician took the active lead. This can be explained asan extension of the physicians traditional legitimacy in the medical status hierarchy, whereasadministrators are a recent addition and have no clear role. But these particular physicians,rather than playing the traditional physician role, used their prestige to deliberately reconstructrelations.

They did not abdicate authority: both were very clear about the need for strong leadership. Atthe same time, however, both leaders unusually and deliberately encouraged nurses to be activein problem-solving. The Hightown leader symbolized this most clearly: it was these AdvancedPractice Nurses who were at the center of the interdisciplinary rounds, while the physician satsomewhat to the side actively participating, but not the focus of attention. The APNs alsooffered advice to outpatient nurses and others who had contact with patients; they were able to

integrate the lengthy course of heart failure treatment better than a physician could, and herecognized the value of this role.

Administrators, as noted before, were less visible. But their role were important: to start with,they hired the key physicians and they provided resources and space for them to operate. Wedid not focus on the administrative role in detail, but it is likely that there is much to be learnedfrom cases like this about how administration can provide a framework for collaborativerelations without creating the kind of conflict or rejection we saw at Riverside and Lowell.

It is worth repeating, therefore, that at United, the leader worked extremely hard atnetworking with other parts of the hospital who cared for cardiac patients, in order to create auniform standard of care even in areas outside his direct domain. It is possible that thisnetworking model of change might avoid some of the dangers of the consolidating model.

The integration of cost / efficiency values with quality of care

We initially expected successful collaboration to be based on the single shared value of patientcare which unites all the health professions, and for the most part to reject efficiency or costvalues. But this turned out to be at least partially wrong. In fact we saw this one-dimensionalfocus most clearly in the two unsuccessful hospitals. Many of the care staff, like the nurse atRiverside quoted above (My mission is caring for the patient, and get out of my way!),expressed moral objections to taking costs into consideration. The administrators at Riverside,meanwhile, dodged the issue by arguing that if they focused on quality it would automaticallylead to a reduction in costs. This avoids dealing with the very real tensions that often crop upbetween the two values; and indeed we did not see much active awareness of or attention to

cost reduction.

But in Hightowns collaborative effort in particular there seemed to be substantial progresstowards integrating the two priorities. Hightown cardiac practitioners sometimes resisted costpressures which they saw as illegitimately compromising quality of care: but they also spoke ofthe other side of the equation, which we did not find elsewhere: they told other stories of active,committed efforts by teams of care providers to attack inefficiencies. This suggests that they


20/30

20

were indeed operating in two dimensions seeking to balancethe values of quality andefficiency, and to work through the contradictions in the workplace as they arose.

We did not get enough evidence on this issue to fully understand how this had happened andhow extensive these attitudes really were. The other collaborative effort, at United, had alsosucceeded in reducing costs, but our interviews turned up more mixed feelings, with some

nurses following the traditional line that this is not something that caregivers should worryabout. This is something worth focusing on in future investigations, since the long-term successof the health system almost certainly depends on building this balancing of competing prioritiesinto daily decision-making at all levels. We have not come across, in our research, convincingevidence of this integration in other settings, though it is unlikely that we stumbled across theonly such case.

Conclusion: The long path forward

This study leads to two broad conclusions for practice. First, the frequent strategy of seekinggreater administrative control over professional practice, in order to force attention to costsand reliability, may lead to a clash of orientations which produces worse results, especially in

complex syndromes, like cardiac care, which are among the costliest and most difficult areas ofmedicine today.

Second, our findings offer optimism but heavily qualified about a change strategy ofbuilding collaboration through physician leadership, building a culture and mechanisms ofshared problem-solving that cut across traditional status and organizational boundaries. Thisstrategy, which we observed in two cases, was more promising than we expected in combiningthe values of quality care with improved cost and efficiency.

We could hypothesize based on our results is that the very best cardiac units will necessarilyoperate according to collaborative principles. That was true in our small sample, and we couldsee in detail dynamics that enabled our two collaborative cases to deal exceptionally effectively

with the complexity of heart failure. Nevertheless, it would of course take a great deal moreresearch, including close study of possible exceptions, before this rule could be convincinglyestablished.

The results also show the significance of a cultural and organizational view of the problem.Significant improvement in medical care delivery involves far more than installing newmechanisms or incentives; it requires reconfiguration of expectations and orientationsthroughout the organization. Many of the changes affect basic approaches to medicine that dateback more than a millennium and that permeate the entire medical complex.

Thus even if the claim about the power of collaboration is true, there are clearly many obstaclesto generalizing it as a change strategy. Existing institutions professional bodies and norms,

other units of hospitals and the wider care system, and the insurance reimbursement system create a large number of boundary problems that can easily lead to isolation and loss ofmomentum.

This analysis does not suggest actions that can easily or rapidly improve health care delivery. Ifanything, it underlines the magnitude of the challenge and implies that it will involve a longchange process. The evolution of collaborative systems in private sector businesses has beengoing on for 20 to 30 years and is far from complete: it involves extensive changes not only inmechanisms, which can be modified relatively easily, but also in mindsets and relationships,


21/30

21

which are far more difficult. Health care, as we can see from the stories above, involves at leastas difficult a transformation in attitudes.


22/30

References

Adler, Paul S., and Bryan Borys. 1996a. Two Types of Bureaucracy: Enabling and Coercive.Administrative Science Quarterly41 (1).

Adler, Paul S., and Charles Heckscher. 2006. Towards collaborative community. In The Firm as aCollaborative Community: Reconstructing Trust in the Knowledge Economy, ed. CharlesHeckscher and Paul S. Adler, 355-416. Oxford, England: Oxford University Press.

Adler, Paul S., S.-W. Kwon, and Charles Heckscher. 2008b. Professional Work: The Emergenceof Collaborative Community. Organization Science19 (2): 359-376.

Alt-White, A. C., M Charns, and R. Strayer. 1983. Personal, organizational and managerialfactors related to nurse-physician collaboration. Nurs Adm Q8 (1): 8-18.

Anon. 2005. Doctors in Society: Medical Professionalism in a Changing World. London: Royal

College of Physicians.

Argote, L. 1982. Uncertainty and Organizational Coordination in Hospital Emergency Units.Administrative Science Quaterly27: 402-434.

Arndt, M., and B. Bigelow. 2005. Professionalizing and masculinizing a female occupation: thereconceptualization of hospital administration in the early 1900s.Administrative ScienceQuarterly50 (2): 233261.

Asch, S. M., D. W. Baker, J. W. Keesey, M. Broder, M. Schonlau, M Rosen, P. L. Wallace, and E.B. Keeler. 2005. Does the Collaborative Model Improve Care for Chronic HeartFailure?Med Care43 (7): 667-675.

Berg, Peter, Eileen Appelbaum, Thomas Bailey, and Arne L Kalleberg. 1996. The PerformanceEffects of Modular Production in the Apparel Industry. Industrial relations.35 (3): 356.

Berry, L. L. 2004. Leadership lessons from Mayo Clinic. Organizational Dynamics33: 228242.

Carmel, S. 2006. Boundaries obscured and boundaries reinforced: incorporation as a strategy ofoccupational enhancement for intensive care. Sociology of health & illness(Print)28 (2):154-177.

Chandler, A. D. 1977. The Visible Hand: The Managerial Revolution in American Business. HarvardUniversity Press.

DAmour, D., M.Ferrada-Videla, L. San Martin Rodriguez, and M. D. Beaulieu. 2005. Theconceptual basis for interprofessional collaboration: core concepts and theoreticalframeworks.Journal of Interprofessional Care19 (Suppl 1): 116-31.

DAmour, D., L. Goulet, R. Pineault, J. Ladadie, and M. Remondin. 2004. Comparative study ofinterorganizational collaboration and its effects in four Quebec health regions: The case


23/30

23

of perinatal services.Montral: Groupe de recherche interdisciplinaire en sant (GRIS),Universit de Montral.

DAmour, D., C. Sicotte, and R. Lvy. 1999. Collective action within interprofessional teams inhealth services. Sciences Sociales et Sant17 (3): 67.

Daiski, I. 2004. Changing nurses dis-empowering relationship patterns.Journal of AdvancedNursing48 (1): 43-50.

Davis, K., C. Schoen, and K. Stremikis. 2010. How the Performance of the U.S. Health Care SystemCompares Internationally 2010 Update. The Commonwealth Fund, June.

Deal, T. E., and A. A. Kennedy. 2000. Corporate cultures: The rites and rituals of corporate life. BasicBooks.

Ducharme, A., O. Doyon, M. White, J. L. Rouleau, and J. M. Brophy. 2005. Impact of care at amultidisciplinary congestive heart failure clinic: a randomized trial. Canadian Medical

Association Journal173 (1): 40-45.

Edmondson, A. C. 2003. Speaking Up in the Operating Room: How Team Leaders PromoteLearning in Interdisciplinary Action Teams.Journal of Management Studies40 (6): 1419-1452.

. 2004. Learning from failure in health care: frequent opportunities, pervasive barriers.British Medical Journal13 (suppl_2).

Engel, G. V. 1970. Professional Autonomy and Bureaucratic Organization.AdministrativeScience Quarterly15 (1): 12-21.

Estabrooks, C. A., W. K. Midodzi, G. G. Cummings, K. L. Ricker, and P . Giovannetti. 2005. Theimpact of hospital nursing characteristics on 30-day mortality. Nursing Research54 (2):74-84.

Feldman, M. S, and B. T Pentland. 2003. Reconceptualizing Organizational Routines as aSource of Flexibility and Change.Administrative Science Quarterly48 (1): 94121.

Freidson, E. 2001. Professionalism: The Third Logic (Cambridge: Polity). Links.

Galbraith, J.R. 1994. Competing with Flexible Lateral Organizations. 2nd ed. Reading, Mass:Addison-Wesley.

Gawande, Atul. 2007. Annals of Medicine: The Checklist: Reporting & Essays: The NewYorker. The New Yorker, December 10.http://www.newyorker.com/reporting/2007/12/10/071210fa_fact_gawande?printable=true.

Gibbon, B. 1999. An investigation of interprofessional collaboration in stroke rehabilitationteam conferences.Journal of Clinical Nursing8 (3): 246-252.


24/30

24

Gittell, J. H. 2000. Organizing work to support relational co-ordination. International Journal ofHuman Resource Management11 (3): 517-539.

. 2002. Coordinating mechanisms in care provider groups: relational coordination as amediator and input uncertainty as a moderator of performance effects.ManagementScience48 (11): 1408-1426.

Gittell, J. H., K. M. Fairfield, B. Bierbaum, W. Head, R. Jackson, M. Kelly, R. Laskin, S. Lipson, J.Siliski, and T. Thornhill. 2000. Impact of Relational Coordination on Quality of Care,Postoperative Pain and Functioning, and Length of Stay: A Nine-Hospital Study ofSurgical Patients.Medical Care38 (8): 807.

Goode, W. J. 1957. Community within a community: the professions.American SociologicalReview22 (2): 194-200.

Goodman, P. S. 1980. Realities of improving the quality of work life: Quality of Work Lifeprojects in the 1980s. Labor Law Journal31 (8): 487-494.

Grant, R. M. 1996. Toward a knowledge-based theory of the firm. Strategic Management Journal17 (10): 109-22.

Gugan, Luc. 1995. Hpital, innovation, professions. CFDT Aujourdhui, June.

Hall, Pippa. 2005. Interprofessional teamwork: Professional cultures as barriers.Journal ofInterprofessional Care19 (2 supp 1): 188. doi:10.1080/13561820500081745.

Heckscher, Charles. 2007. Types of strategy. In The Collaborative Enterprise: Managing Speed andComplexity in Knowledge-Based Businesses, ch. 3. Yale University Press.

Hoff, T. J. 1999. The Social Organization of Physician-Managers in a Changing HMO. Work andOccupations, 26(3), 324 -351.

Hyrkas, K., and K. Appelqvist-Schmidlechner. 2003. Team supervision in multiprofessionalteams: team members descriptions of the effects as highlighted by group interviews.Journal of Clinical Nursing12 (2): 188-197.

Kieser, Alfred. 1989. Organizational, Institutional, and Societal Evolution: Medieval CraftGuilds and the Genesis of Formal Organizations.Administrative Science Quarterly34 (4)(December): 540-564.

Krause, E. A. 1999. Death of the Guilds: Professions, States, and the Advance of Capitalism, 1930 to thePresent. Yale Univ Pr.

Leape, L. L., and D. M. Berwick. 2005. Five years after to err is human: what have we learned?Jama293 (19): 2384.

Liedtka, J. M., and E. Whitten. 1998. Enhancing care delivery through cross-disciplinarycollaboration: a case study.Journal Of Healthcare Management43 (2): 185-203.


25/30

25

Light, D., and S. Levine. 1988. The Changing Character of the Medical Profession: ATheoretical Overview. The Milbank Quarterly66: 10-32.

Maccoby, Michael. 2006. Health care organizations as collaborative learning communities. InThe Firm as a Collaborative Community: reconstructing trust in the knowledge economy, ed.Charles Heckscher and Paul S. Adler. Oxford: Oxford University Press.

Makary, M. A., J. B. Sexton, J. A. Freischlag, C. G. Holzmueller, E. A. Millman, L. Rowen, and P.J. Pronovost. 2006. Operating Room Teamwork among Physicians and Nurses:Teamwork in the Eye of the Beholder.Journal of the American College of Surgeons202 (5):746-752.

Mathison, S. 1988. Why Triangulate? Educational Researcher17 (2): 13.

McAlister, F. A., S. Stewart, S. Ferrua, and Jjjv McMurray. 2004. Multidisciplinary strategies forthe management of heart failure patients at high risk for admission A systematic reviewof randomized trials.Journal of the American College of Cardiology44 (4): 810-819.

Merton, R. K. 1940. Bureaucratic Structure and Personality. Social Forces18 (4): 560-568.

Miles, M. B., and A. M. Huberman. 1994. Qualitative Data Analysis: An Expanded Sourcebook. SagePublications.

Montgomery, J. 1989. Medicine, Accountability, and Professionalism.Journal of Law andSociety16 (3): 319-339.

Nonaka, I. 2005. A Dynamic Theory of Organizational Knowledge Creation. KnowledgeManagement: Critical Perspectives on Business and Management5 (1): 14-37.

Parsons, T. 1963. Social change and medical organization in the United States: a sociologicalperspective. The ANNALS of the American Academy of Political and Social Science346 (1):21.

. 2007.American society: a theory of the societal community. Ed. G. Sciortino. ParadigmPublishers.

Powell, W. 1990. Neither Market Nor Hierarchy: Network Forms of Organization. Research inOrganizational Behavior12: 295-336.

Ramanujam, Ranga, and Denise Rousseau. 2006. The challenges are organizational not just

clinical.Journal of Organizational Behavior27 (7): 811-827.

Risse, G. B. 1999.Mending bodies, saving souls: a history of hospitals. Oxford.

Rosen, E. 2007. The Culture of Collaboration. Maximizing Time, Talent and Tools to Create Value inthe Global Economy. Red Ape Publishing, San Francisco.

Rubinstein, Saul A., and Thomas A. Kochan. 2001. Learning from Saturn : possibilities for corporategovernance and employee relations. Ithaca: ILR Press.


26/30

26

Sabel, C. 1991. Moebius-Strip Organizations and Open Labor Markets: Some Consequences ofReintegration of Conception and Execution. In Social Theory for a Changing World, ed.Pierre Bourdieu and James S. Coleman, 2353.

San Martn-Rodrguez, L., M. D. Beaulieu, D. DAmour, and M. Ferrada-Videla. 2005. Thedeterminants of successful collaboration: A review of theoretical and empirical studies.Journal of Interprofessional Care19: 132-147.

Scott, W. R., M. Ruef, P. Mendel, and C. A. Caronna. 2000. Institutional Change and Organizations:Transformation of a Healthcare Field. Chicago: Univ. Chicago Press.

Shojania, K G, B W Duncan, K M McDonald, R M Wachter, and A J Markowitz. 2001. Makinghealth care safer: a critical analysis of patient safety practices. Evidence report/technologyassessment (Summary)(43): i-x, 1-668.

Sicotte, C., D. DAmour, and M. P. Moreault. 2002. Interdisciplinary collaboration withinQuebec community health care centres. Social Science & Medicine55 (6): 991-1003.

Siln-Lipponen, M., H. Turunen, R. N. LicNSc, and K. Tossavainen. 2002. Collabora tion in theOperating Room: The Nurses Perspective.JONA: The Journal of Nursing Administration32 (1): 16.

Spear, S. J. 2005. Fixing healthcare from the inside, today. Harvard Business Review83 (9): 78-91.

Street, R. L., H. Gordon, and P. Haidet. 2007. Physicians communication and perceptions ofpatients: Is it how they look, how they talk, or is it just the doctor? Social Science &Medicine65 (3): 586-598.

Trochim, W. 1989. Outcome pattern matching andprogram theory. Evaluation and ProgramPlanning12 (4): 355-366.

Weber, Max. 1946. Class, Status, and Power. In From Max Weber: Essays in Sociology, ed. HhGerth and Cw Mills. New York: Oxford University Press.

Weinberg, D. B. 2003. Code Green: Money-Driven Hospitals and the Dismantling of Nursing. CornellUniv Pr.

Wensing, Michel, Peter Vedsted, Janko Kersnik, Wim Peersman, Anja Klingberg, HilaryHearnshaw, Per Hjortdahl, et al. 2002. Patient satisfaction with availability of general

practice: an international comparison. Int J Qual Health Care14 (2) (April 1): 111-118.

Wotton, K., S. Borbasi, and M. Redden. 2005. When All Else Has Failed: Nurses Perception ofFactors Influencing Palliative Care for Patients With End-Stage Heart Failure.Journal ofCardiovascular Nursing20 (1): 18.

Yin, R. K. 2003. Case Study Research: Design and Methods. Sage Publications Inc.


27/30

27

Young, D, Bonnie Koenig, Adil Najam, and Julie Fisher. 1999. Strategy and Structure inManaging Global Associations. Voluntas: International Journal of Voluntary and NonprofitOrganizations10 (4) (December 1): 323-343. doi:10.1023/A:1021434424549.

Young, GJ, MP Charns, J. Daley, M. G. Forbes, W. Henderson, and S. F. Khuri. 1997. BestPractices for Managing Surgical Services: The Role of Coordination. Health CareManagement Review22 (4): 72.


28/30

Tables

Table 1: three models of organizational culture

Traditional craft /

autonomous professional

Administrative /

bureaucratic

Collaborative

Key

characteristics

Norm of individual

autonomy for lead

practitioner

Accountability to

professional organization or

guild

Formal organization small

or absent

Strong status hierarchy

(usually): Personal relationto clients

Strong hierarchical

organization

Strong boundaries

Jobs defined in relation

to organization

structure

Power based on

position

Resources linked

directly toaccountability

Two dimensions:

hierarchical control &

systematic processes

Porous boundaries

Rapid reconfiguration

around problems /

processes

Influence based on

knowledge / contribution

rather than status or

position

Differentiation of resources

from accountability

General business

history

Dominant from medieval

period through 19thcentury

Rise of large firms at

end of 19thcentury,

consolidated by mid-

20th

Gradual emergence of

complex teams and process

organization from mid-

1970s, still incomplete

Health care

history

Dominant model until

recently, now under

pressure from specializationand cost increases

Rapid growth of

administrative control

mechanisms (insurance,HMOs, etc) from

1980s, has not

displaced craft

organization at patient

interface

Much recent

experimentation, little

clarity of vision or practice

Table 2: Distribution of cases(pseudonyms are used throughout for the hospitals)

rich payer mix poor payer mix

High performing Hightown United

Low performing Riverside Lowell


29/30

29

Appendix

Table 3

Description of Case Data

Hospital/Characteristics

Riverside Lowell Hightown United

Description Medium sized(300 beds?)

300 beds Large communityhospital

Large teachinghospital

Cultural system Craft-based(traditionalprofessionalculture), mildefforts atadministrativereforms

Administrativereforms in conflictwith traditionalprofessionalculture in

Limitedcollaboration inwealthy setting,relativelydeveloped

Limitedcollaboration incost-constrainedsetting, relativelynew

Performance Bottom quarterof state on bothadjusted in-hospital mortalityand processconsistency

Bottom quarter ofstate on bothadjusted in-hospital mortalityand processconsistency

Top quarter ofstate on bothadjusted in-hospital mortalityand processconsistency

Top quarter ofstate on bothadjusted in-hospital mortalityand processconsistency

Wealth of patients(payer Mix)

Wealthy payermix

Poor payer mix Wealthy payer mix Poor payer mix

Total interviews 27 28 25 25

AdministratorsPhysiciansHome health

careNurses and

nursemanagers

Dietician /nutritionist

PharmacistsHome care /

outpatientSocial worker

Clinical researchCase managersUnit secretaries

and clerksPatient

advocatesTransporter

AdministratorsPhysicians

(includinghospitalists)

Home health careNurses (agency

and staff; nursemanager)

Dietician /nutritionist

PharmacistHome care /

outpatient

Social workerClinical researchCase managersUnit secretaryPatient advocates

AdministratorsPhysiciansHome health careNurses (director of

nursing;outpatient,clinical)

Medicaltechnologist

DieticianPharmacistHome care /

outpatient nurses

Social workerClinical research

AdministratorChief of

cardiology2 physician

fellowsPhysiciansNurses (telemetry,

qualityassurance, staff,APN, setc)

Medicaltechnologists

Pharmacist

Outpatientpsychologist

Case managerPhysical

therapists /dieticians


30/30

1For the purposes of this paper we are conflating craft and professional cultures. They aresimilar on the essential organizational dimensions: the key unit is the independent holder ofspecialized knowledge; if others are involved, it is as strongly deferential subordinates withlower status (e.g., apprentices or helpers such as traditional nurses).

2The American Heart Association reports that coronary heart disease caused 451,326 deaths in2004 and is the single leading cause of death in America today.(http://www.americanheart.org/presenter.jhtml?identifier=4478,4/2/08). In that year heartfailure was a contributing factor in over 284,000 deaths; the estimated direct and indirect costof HF in the United States for 2008 is $34.8 billion.
http://www.americanheart.org/presenter.jhtml?identifier=4478http://www.americanheart.org/presenter.jhtml?identifier=4478http://www.americanheart.org/presenter.jhtml?identifier=4478http://www.americanheart.org/presenter.jhtml?identifier=4478

-collaboration and the quality of health care delivery

Documents