Characteristics of the participants and setting
The study took place at a Canadian province-wide health call centre serving a population of 4.2 million and averaging 22,600 calls per month. Of the 31 nurses who participated in the implementation intervention, 25 nurses (80.6%) responded to the survey and 8 participated in the focus groups (Table
1). The subset of nurses who completed the survey shared similar demographic characteristics with all those who participated in the intervention (Table
2). Four administrators (key informants) were also interviewed.
Table 2
Characteristics of the participants by data source
Frontline staff nurses | 8 (66.7) | 25 (100) | 31 (100) |
Nurse supervisor or educator | 2 (16.7) | 0 | 0 |
Non-nurse administrators | 2 (16.7 | 0 | 0 |
Length of employment | | | |
≤ 6 months | 1 (8.3) | 4 (16.0) | 5 (16.1) |
7 to 12 months | 3 (25.0) | 8 (32.0) | 9 (29.0) |
>12 months | 8 (66.7) | 13 (52.0) | 17 (54.8) |
Employment status (full-time equivalent) | Mean 0.77 | Mean 0.75 | Mean 0.74 |
not reported (casual status) | 1 (8.3) | 2 (8.0) | 2 (6.5) |
BSc or higher education | 7 (58.3) | 10 (40.0) | 13 (41.9) |
Gender | | | |
Female | 10 (83.3) | 25 (100) | 30 (96.8) |
Male | 2 (16.7) | 0 | 1 (3.2) |
Years of nursing | | | |
≤ 5 years | 0 | 0 | 0 |
6 to 10 years | 0 | 1 (4.0) | 2 (6.5) |
11 to 15 years | 1 (8.3) | 7 (28.0) | 7 (22.6) |
≥ 16 years | 9 (75.0) | 17 (68.0) | 22 (71.0) |
not reported | 2 (16.7) | 0 | 0 |
N
| 12 | 25 | 31 |
The call centre provides toll-free 24-hour telephone consultation by registered nurses to help residents manage their health and participate actively in making health decisions. Unique from other call centres in Canada [
9], this call centre is part of an integrated self-care program that also provides public access to a self-care handbook and Internet-based health information resources, including over 95 patient decision aids. Monthly reports, from December 2003 to June 2004, indicated that about 55% of the calls concerned triaging symptoms, 25% were about a specific health condition, and 20% concerned other issues (e.g., drug information, finding health services). In 2003, the most common patient decision aids accessed by the nurses included those dealing with birth control methods, breast versus bottle feeding, male newborn circumcision, wisdom teeth removal, and treatment of miscarriage.
This call centre was established in April 2001 when the provincial ministry of health awarded a three-year contract to a private, not-for-profit management company. Of the 108 nurses employed in 2003, the typical nurse was female, had over 20 years of nursing experience, worked part-time hours, was unionized, and had worked at the call centre for one year or longer [
12]. Nurses were grouped into three teams, each led by a nursing supervisor who reported directly to a non-nurse operations manager and indirectly to a nursing practice leader. The operations manager and nursing practice leader reported to a director of operations who reported to the provincial ministry of health.
Several mechanisms were in place to ensure program quality and minimize the risk of litigation [
19,
20]. On hiring, nurses received 105 hours of orientation and three months of mentoring. The orientation was focused mainly on triaging symptoms, with 0.75 hours devoted to introducing patient decision aids. The computerized protocols and health information database used to guide the telephone consultations were purchased from Healthwise
® Inc. and adapted for Canadian use. Call centre activities were monitored and reported to the provincial ministry of health on a monthly basis using a set of performance indicators (e.g., respond to 80% of calls within 20 seconds) based on the American Health Call Centre Accreditation Standards of the Utilization Review Accreditation Commission Inc. [
21]. Monthly reports included statistics on call volume, call response time, call abandonment, length of calls, proportion of first time callers, call disposition (e.g., emergency, physician visit, self-care), pre- and post-call intent of the caller, and results of a quality audit on a random sample of audio-taped calls.
Over the study period from December 2003 to June 2004, there were several concurrent activities that were likely to have influenced this implementation study. In January 2004, nursing supervisors' roles and responsibilities were restructured. These changes resulted in the creation of a master staff development plan, major change in staffing patterns, and an expansion of call centre services (e.g., palliative care, newborn care). In March 2004, nurses started verifying caller demographics by linking to the provincial ministry of health's confidential database. Implementation of this practice change involved classroom training of all staff and subsequent performance review by nursing supervisors on real calls prior to autonomous practice. Over the study period, nurse absenteeism and inadequate staffing resulted in a higher number of calls in the hold queue and as a result, increased pressure for nurses to shorten their call length. Finally, the contract for the call centre services was due for renewal in the summer of 2004, which caused concern about job security among the nurses, increased organizational pressure to meet performance indicators, and re-directed administrative priorities to preparing a response to the imminently expected request for proposals.
Was the decision support protocol adopted into clinical practice?
Eleven of the 25 nurses (44%) who participated in the implementation intervention (e.g., autotutorial, skill building workshop, decision support protocol, performance feedback) used the decision support protocol within one month following the intervention and the remaining 14 nurses (56%) reported that they had not received calls requiring values-sensitive decision support during that time. Twenty-one nurses (84%) agreed that they were comfortable using the decision support protocol. Most nurses (92%) indicated that they intended to use the protocol within the next three months. Nurses in the focus groups shared their experiences using the decision support protocol with real callers. One nurse spoke of the challenges of getting started and learning through her early experiences.
It was just plunge in, see what you do the first time. So the first few I did, I did all in one day. And I may not have been right on all of them but I could see where I missed. The next one I thought was better.
What effect did the intervention have on nurses' approach to supporting real callers making values-sensitive decisions?
Recognize need for decision support
Nurses reported being more likely to recognize callers experiencing decisional conflict and highlighted the issues related to call classification. One nurse shared, "Whereas before I might have asked a series of questions before I came to a realization that they were in a complex decision making process. Now I can identify much more readily." Nurses identified that these calls would be difficult to identify in the database because they would usually be classified as a health condition-specific or medication-related call.
Improve decision support
Many nurses shared examples of how they thought their approach to providing decision support had improved. To exemplify, one nurse described how the protocol facilitated a more specific assessment: "I'm more likely to ask questions about the decision and where they are on it instead of just making assumptions; which is a lot of what I did earlier." Of the 25 nurses who completed the survey, over 90% agreed that the decision support protocol was logical (n = 23), helped prepare callers for discussing decisions with their practitioners (n = 24), complemented the nurses' usual approach (n = 23), and helped them to more fully explore the issues of importance to the callers (n = 24). All 25 nurses (100%) agreed that the new protocol facilitated caller empowerment. This was further supported by focus group nurses' description of callers being more engaged in the discussion; "...and it's a dialogue and they really feel part of the dialogue." Another nurse shared, "...especially when you ask them the pros and the cons. You know suddenly the light goes on; like, I guess I could write them down." Of the 25 nurses, 24 (96%) agreed that using the protocol provided a more consistent approach to supporting the callers. Several nurses described the new approach for handling these decision support calls as more efficient, streamlined, and shorter (e.g., "with the specific tool to ask, I find the call goes quicker").
Perceived practice changes positively
The importance of nurses providing patient decision support was supported by one nurse who said, "Anybody can read the information the value of nursing in my philosophy is that you're helping counsel, guide. Provide information, yes, but not just a telephone operator." Nurses also appreciated having a structured process for approaching these types of calls which took the pressure off having to find the 'right decision'. For example, "I used to feel quite nervous that I felt like I should know the answer. So this has given me a lot of power that you can help them, that you don't have to sort it out for them." Several nurses expressed their general satisfaction with their enhanced decision support role: "For me, this is the most enjoyable part"; "I came out knowing I made a difference"; "That's the job I want to do, help people making any decisions." Finally, some nurses reflected upon how these new skills in exploring values were relevant to symptom calls, particularly when callers did not agree with the triage decision determined by the protocol. For example, one nurse shared, "when we are sure they should call 911 and they're really reluctant and I always say well, what's the reason behind this so you kind of try to explore."
What factors are likely to influence the sustainability of values-sensitive decision support by call centre nurses?
Barriers and facilitators influencing sustainability are presented in Table
3.
Table 3
Suggestions to enhance sustainability by overcoming barriers to nurses providing values-sensitive decision support
Innovation: Decision Support
| Patient decision aids are hard to use with patients over telephone | - Decision aids need more point form and auto-charting |
| No structured process for preparing callers for shared decision making | - Resolved with use of Decision support protocol. |
| Decision support protocol is not integrated with charting | - Integrate protocol in computer database with auto-charting ability |
Potential Adopters: Nurses
| Inadequate decision support knowledge | - Resolved by providing nurses with access to an autotutorial |
| Inadequate skills in providing decision support | - Partially resolved with nurses participation in skill building workshop - Mentoring from supervisors to further develop nurses' skills - Revise call audit tool to include key decision support elements - Continuing education to reinforce learning - Encourage nurses to self-assess their performance |
| Low confidence in ability to provide decision support | - Nurse supervisors could give positive feedback on quality of decision support provided |
Practice Environment: Call Centre
| Unclear program direction to provide decision support | - Determine impact of decision support calls on program services - Establish clear direction |
| Limited orientation of new staff to decision support resources | - Use feedback to revise implementation intervention - Extend training to all nurses and in-particular nurse supervisors - Revise call audit tool to include elements of quality decision support |
| Pressures to minimize call length | - Revise call classification to collect decision support calls statistics - Establish call length guidelines tailored to types of calls - Revise patient decision aids for easier use by telephone - Integrate decision support protocol into the database |
| Low caller awareness that call centre nurses provide decision support | - Market decision support services to public & other health services |
To facilitate use of the decision support protocol and patient decision aids, nurses need to have these tools readily accessible for use over the telephone. In the survey, 18 nurses (72%) agreed that the protocol, in its current format as a word processing file, takes extra time to navigate, transfer into the documentation system, and use for documenting. By including the protocol as a screen within the documentation system, one nurse suggested, "...as soon as you recognize that somebody is in one of these situations and you can push a button on your screen and have it pop in your call manager. How easy would that be? That would be swell!" Nurses suggested that patient decision aids in the database needed to be easier to locate and revised for use over the telephone. For example, one nurse stated,
And setting it up with pro's, con's, not big sentences to explain each point. I mean if we're supposed to know it, we're supposed to know it. So you know, you might want to have preambles for all this stuff, if you have to. But it is cut and dry. Get it short. Point form.
As well, nurses wanted the protocol and the patient decision aids linked into the documentation system such that nurses' responses to questions would be automatically transferred into the electronic health record; similar to the way in which auto-charting occurs in symptom protocols.
Ongoing reinforcement for skill development
Nurses in the focus groups requested opportunities to support their applying these novel skills in practice. For example, "it would be great just to have more of those simulated calls just to be able to do them" and routine inservices focused on sharing experiences from decision support calls to offer a "feeling of connection with other people who are doing them". The nursing supervisors were identified as those best positioned to mentor the nurses, given their current responsibilities include providing feedback from call audits and coaching nurses to improve call handling. One nurse shared, "If there is a problem with your times, what she [nursing supervisor] does is goes over that with you and tries to coach you and pulls calls that are long to see, you know, where you need shortening." Nurses also expressed concern about a patient decision support call being randomly selected for the monthly call audit. "I don't think that they [nursing supervisors] would know how to acknowledge what was done well and try to coach to what other things could be done better." Although nursing supervisors were invited to participate in the intervention as non-study participants, competing demands due to organizational changes limited their availability to participate. Furthermore, their call audit tool did not include key elements necessary for quality patient decision support.
Fit of decision support with program direction
Nurses suggested that if supporting callers facing values-sensitive decisions is an expectation of their role this needed to be made clear in the program direction. To that end, appropriate changes would need to be made to organizational policies and procedures. Nine of 25 nurses (36%) felt that they had clear direction from the organization that they should be providing values-sensitive decision support. One administrator appeared less sure about the need for an organizational directive specific to providing patient decision support; "...to communicate the value that this is a positive change for nursing practice as it takes the [call centre] in a new direction, in a direction I think we want to go in". Administrator key informants identified that prior to an organizational commitment to having call centre nurses provide values-sensitive decision support they needed to determine the impact on call centre staffing, performance monitoring, the nursing education plan, and budget. Most nurses agreed that the call centre services should include patient decision support with 20 (80%) identifying all nurses and 2 (8%) identifying only a sub-group of specialized nurses as those who should be providing decision support guided by the new protocol.
Decision support training for other nurses
Of those surveyed, 22 nurses agreed (88%) that nurses would need education sessions, beyond their initial call centre orientation, to develop their knowledge and skills in values-sensitive decision support. In one focus group, a nurse suggested the need to "...
embed it in our continuing education program". Our implementation intervention (i.e., decision support protocol, autotutorial, workshop, and performance feedback on simulated calls) was acceptable to over 90% of participants [
13] and could be used for ongoing decision support training. Focus group nurses offered to be simulated patients for other staff developing these skills. The best timing for this type of training ranged from "
in their orientation week or the week after their orientation week so that they start out doing this when they're taking decision making type calls" to three to six months after starting at the call centre.
Call length guidelines
Throughout the study, nurses were concerned that decision support calls would take longer than the organizational 12.5 minute call length target and requested that call length guidelines be tailored to types of calls. Despite this frequently identified barrier, one nurse in the focus group shared how she rationalized longer calls,
...so I personally don't worry about it. And I find it all balances out...If you don't deal with it now, then it sort of goes down the line. It's going to take more time and money and everything else.
Nurses also highlighted environmental pressures to minimize call length that included the flashing light on their telephone to indicate waiting calls, an electronic display board to indicate the number of callers waiting, and personal monthly reports on call length.
Marketing of decision support services
Administrators and nurses argued that, for sustainability, the public and health care providers needed to be informed about the decision support services available through the call centre. This was supported by the survey finding that only 4 of 25 nurses (16%) agreed that the public was aware that the call centre nurses could support people facing values-sensitive health decisions.