Information Needs and the Use of Documentation to Support Collaborative Decision-Making

Abstract

It is clear that interdisciplinary communication and collaboration have the potential to mitigate healthcare-associated harm, yet there is limited research on how communication through documentation in the patient record can support collaborative decision making. Understanding what information is needed to support collaborative decision making is necessary to design electronic health information systems that facilitate effective communication and, ultimately, safe care. To explore this issue, we focused on information needs related to central venous catheter management and the prevention of central line–associated blood stream infections. Semistructured interviews were conducted with nurses working in an intensive care unit. Interview transcripts were analyzed using inductive thematic analysis. Three themes were identified: (1) challenges managing documentation in multiple places in the absence of formal documentation processes for central venous catheter management; (2) lack of standardized decision-making processes for managing central venous catheters; and (3) oral communication holds it together. Our findings provide a foundation for the development of EHR functional requirements that enhance communication regarding the management of central venous catheters and facilitate the prompt removal of unnecessary lines.

Documentation in the electronic health record (EHR) serves several purposes; however, a primary and often overlooked purpose is facilitating communication among the healthcare team. Communication failures are one of the most ubiquitous contributors to healthcare-related harm. With an estimated 210 000 to 400 000 deaths per year, healthcare-associated harm is equivalent to the third leading cause of death in the United States.^1,2 It is also clear that effective interdisciplinary communication and collaboration are necessary for the provision of safe care.^3,4 The shift from paper-based records to computerized systems or EHRs has affected how documentation is used as a means for communication among the healthcare team. The ubiquitous nature of data in the EHR has the potential to improve information sharing among all users. Yet, there is a need for more research on how clinical documentation in the EHR is used in practice to meet interprofessional communication needs.

Although there has been limited research on how documentation supports communication, there is some evidence to suggest that clinicians demonstrate a preference for oral communication⁵ and that documentation is not regularly used as a means of communication, particularly among members of different professional groups.^6,7 This could be due to the perception that documentation is not up-to-date^5,8 and that it lacks key information from oral exchanges.^5,9,10 Other researchers have found that clinicians report difficulties in finding key information in the electronic record,^5,8,11 have concerns regarding the low-quality documentation in part due to constraints in computerized systems,^8,10 and believe that notes are not read by fellow clinicians,^10,11 further dis-couraging the use of documentation for information sharing. A few studies have explored the patterns of use of documentation for communication,^5,8–12 yet further research is needed to identify what, where, and how information is documented and shared among the team to better support communication tasks. Identifying best practices related to documentation and implementing these best practices will be critical to facilitating improved communication via the EHR, particularly when face-to-face communication is limited. This study focuses specifically on communication needs to support the prevention of central line–associated blood stream infection (CLABSI).

Central line–associated blood stream infection was selected due to evidence that demonstrates that a reduction in infections is amenable to intervention^13–17 and the ongoing efforts to achieve zero hospital-acquired infections.^15,18,19 Achieving a rate of zero or sustained decreases in CLABSI remains a challenge. This may be due to a disproportionate focus on bundled practices at insertion as compared to the length of time the line is in place, the maintenance of the central venous catheter (CVC), and the access practices.^18–24 These issues, reducing CVC dwell time and adherence to standard maintenance and access practices, have the potential to be improved through EHRs that support information needs related to CVC management. Research has shown that in 21.2% of cases, providers were unaware that one of their patients had a CVC in place,²⁵ and in 8.3% of cases, nurses and physicians were unaware of the rationale for the CVC.²⁶ Another study found that 26.2% of CVCs remained in place with no medical indication after the patient was discharged from the ICU.²⁷ While recent reports demonstrate some progress in the reduction of CLABSI, particularly in the ICU, progress in non-ICU settings continues to lag.²⁸ The 2016 national target was to achieve a 50% reduction in CLABSI by 2020. The actual overall reduction for 2017 to 2018 was 9%.²⁹ As we strive to eliminate these preventable hospital-acquired infections, understanding how to standardize information sharing based on best practices remains relevant. The purpose of this study was to identify what information is used by nurses to determine whether to keep or remove a patient’s CVC, where this information is found, and how this information is shared with the care team.

METHODS

Setting

This qualitative descriptive study took place in the medical intensive care unit of a large academic teaching hospital that used a hybrid paper and electronic documentation system. Interviews took place on the unit, outside the nurse’s assigned rooms, allowing nurses to reference all data sources used to manage patient CVCs. Data collection for this study occurred just prior to the implementation of an enterprise-wide vendor-based EHR; as such, a variety of documentation processes were in use. The clinical documentation workflow has been described in detail by Collins and colleagues.³⁰ Nurses were using paper flowsheets to document daily assessments, intake and output, and vital signs. Medications, provider orders, and laboratory results were accessed and documented using an electronic system. In addition, the unit had recently created an electronic safety checklist, to be used to prompt discussions in interprofessional rounds. The checklist was in development and allowed providers (physicians) to enter data, but it did not automatically pull data from the EHR.³¹

Participants

Semistructured interviews were conducted with a convenience sample of eight registered nurses. Nurses from the ICU were selected due to their regularity in caring for patients with a CVC and, thus, their expertise regarding information needed to support CVC maintenance and access practices, as well as decisions to keep or remove a central line. Institutional review board approval was obtained prior to conducting the study and informed consent was obtained from all participants. Of the eight participants, their experience working as a nurse ranged from 6 to 30 plus years in practice. Eighty-eight percent of the participants interviewed were female, and all but one participant worked more than 31 hours per week.

Interviews

The interviews were conducted opportunistically during the nurses’ shifts and took place just outside their assigned patients’ rooms. As a result, the participants were able to refer to documentation paperwork and tools, as well as to data or cues within the patient’s room, such as dates noted on the patient’s CVC dressing. Two researchers (JT, BC) were present for all of the interviews and individually took notes during each interview. While participants naturally referenced specific documentation tools, the aim of the interviews was to understand information needs based on the nurse’s clinical expertise regarding CVC management and CLABSI prevention, rather than the merits of a particular form or tool.

The semistructured interview guide comprised questions compiled after a review of the literature on CVC management and CVC care bundles. The interview questions explored the nurse’s role in decisions regarding CVC access and removal, what data supported the need to keep or discontinue a CVC, where that data were currently captured or recorded, and how that data or information was communicated among the healthcare team. Additionally, nurses were asked about catheter maintenance and care, including where this information was documented in the record. Minor adjustments were iteratively made to the interview questions throughout the data collection period based on insights gained from interviews. Prompts, such as “what about frequent blood draws” or “how do you know when a dressing change is needed,” were used to encourage nurses to think about all of the different data in the record related to CVC use and management.

Data Analysis

Data were analyzed using inductive thematic analysis. Audiotaped interviews were transcribed verbatim and verified for accuracy by one researcher. Transcribed files were imported into NVivo’s qualitative analysis software (QSR International, Doncaster, Australia). An iterative process of reading and rereading data continued until preliminary themes emerged and all relevant data were categorized. This process moved between individual analysis and group discussion until the final themes and subthemes were agreed upon. In accordance with qualitative design, data collection continued until saturation was reached.³² Although data saturation is a debated concept, in this study, the small sample size proved adequate due to the specificity of the sample, their knowledge of the topic, the use of preexisting standards related to CVC management and CLABSI (established theory), and the focused aim of the study.³³

RESULTS

Three themes were identified: (1) challenges managing documentation in multiple places in the absence of formal documentation processes for CVC management; (2) lack of standardized decision-making processes for managing CVCs; and (3) oral communication holds it together.

Theme 1: Challenges Managing Documentation in Multiple Places

This theme reflected the concept that data related to CVC management were captured in disparate areas of the patient record, as well as in documents that existed outside the patient record. See Table 1 for a summary of themes and subthemes with selected participant quotes supporting the themes. A common theme among participants was the importance of input from the intravenous (IV) therapy nurse who would assess for the viable placement of peripheral IV access, which is necessary prior to removing a CVC; however, this assessment was documented in a separate binder outside the patient record. Another key finding was the reported lack of formal processes for capturing information related to the CVC. Participants remarked that “we have multiple places as nurses to document the one thing,” contributing to double documentation or inconsistencies in where particular information might be found in the record. Data relevant to the management of CVCs was found in disparate areas of the record, including site assessments, dressing changes, insertion date, medication administration records, nursing or synthesis notes, and vital signs documented on the flow sheet. This issue was further complicated by the perception that physicians typically only view certain aspects of nursing documentation, which has been noted in other studies.^6,7 The participants also noted a lack of a specified place for nurses to document the reason for the CVC.

Table 1.

Themes With Subthemes and Supporting Participant Quotations

Subtheme	Quotation
Theme 1: Challenges managing documentation In multiple places
Documentation related to CVC management found in disparate areas of the patient record	“We have multiple places as nurses to document the one thing…we can write it on the flow sheet… and we have an empty space to talk about events over the course of the day….then on our assessment there’s a spot to talk about central lines…and another spot to talk about skin assessments. So if the patient has a red, or irritated, or infected-looking [skin surrounding the CVC]—I’ll write it there [in the skin assessment]. And then there’s always the synthesis notes. So there’s definitely lots of places to—Too many.”
	“It would just be nice to have [insertion sites and dates] all together. So you are not having to put it in five different locations…insertion site, insertion date…dressing change date.
Only certain aspects of nursing documentation viewed by physicians	The back of the flowsheet is where there is documentation regarding the CVC site, but “doctors do not look at the back of the flowsheet.”
	“Definitely the front of the flowsheet [is more useful for communicating with the team]. Very rare that they’ll look at our assessment on the back, and even rarer that they’ll read our synthesis noteand a lot of times I’ll circle something in red if I want todraw the physician’s eyes to it. Because they use…the front of the flow sheets during rounds.”
The patient record is missing information important to CVC management	“Maybe the IV nurse documents multiple attempts to get a PIV [peripheral IV] and the need fora CVC. Nurses might document why a CVC is still needed in the nursing note.”
	The IV nurse documents that a patient has poor access “on the IV therapy sheet” butthis is not in the patient record, it is in “a separate book [binder].”
Theme 2: Lack of standardized decision-making processes
Clinician characteristics influence physician receptiveness to input	“[whether] your opinion is heard is dependent on the physician, you have to make an argument”
	“I guess [the physicians know when I think a patient needs a central line] because I’ve been here 35 years.”
Indications for CVCs differ among clinicians	“[Indicators include] multiple drips, pressors, reliable access, the nurse may want the line for blood sampling, physicians may want the line for pressure monitoring but that is very dependent on the physician and their preference.”
	“[The physician] may want to monitor a CVP, where we are really not thinking it’s all that important. You get your urine output, you get your blood pressure, and what’s that telling you, really? And so, for us, it’s more of an ease of [medication administration], they are on like five or six antibiotics, plus, three or four drips, so you are trying to manage all those things, [with] two or three lines.”
	“I feel like the [physicians] who want them out always follow the data. And then [there are others] who are just like, ‘You know this patient is not going to be with us in the next 48 hours, do not go near the central line. I do not want to cause them any more distress.’ So it’s just different styles.”
	“The other people to get is the IV nurse’s input…because they are first [to say] ‘we cannot do it anymore. We cannot maintain the [peripheral] IV access they are asking for.’”
	“Most often nurses do not know or advocate for a central line to come out, because of the access issues.”
Processes that support decision making	“[The safety checklist is] just so new, we need to get used to it, but [it’s] helpful because it gets the discussion going.”
	“I think we [the nurse] should be able to put why we are using it [the CVC]—can it come out or stay in? The doctors are supposed to be doing that, but I think nurses have a better understanding, and know why it’s in or out. I’m looking at the [checklist] now, which we are not supposed to be doing. It’s the doctor’s eye, and they do not fill it in, so I have gone and filled it in, but we are not supposed to be.”
	“[The nurse’s input is usually sought] on rounds, we [nurses and physicians] decide whether or not they need one [CVC].”
Variations in CVC care	“[for daily care]…we just inspect it [the CVC insertion site], because we have clear dressings on them. I think it’s [every] 96 [hours that they get changed though]—I forget the amount of hours.”
	“if…[there is no] visible blood or any soiling,..they are [dressings are] good for 72 hours. We change the tubing and the claves every 96 hours. We probably [change them] more than we are supposed to. A lot of times that little date sticker [on the dressing] will fall off or something, and then if you do not know, you just change it.”
	“We do not have a great system of knowing when the Luer locks were put on the claves. [They are changed] every 72 hours. [The dressing is changed] weekly, unless soiled. The tubing every four days.”
Theme 3: Oral communication holds it together
Checklist serves as a cognitive aid in the absence of an already identified and discussed need	“We do document the presence of a central line on the [safety checklist] and that cues us to talk about it.”
Oral communication is relied on when a patient condition changes	“If the issue requires immediate attention the nurse will ‘go to them’ [physician], otherwise it is discussed on rounds.”
	“They [physicians] do evening pass offs, so they are [the physicians] checking in on your flowsheets frequently. But most of us, as soon as you have a fever the nurse will go tell the docs.
Proximity of the team promotes oral exchanges	“It could be discussed on admission or many different times during the day, because the team is right there [looking down the unit toward a small office space].”
	“I find that consulting teams read our synthesis notes way more. I cannot ever say I’ve seen like one of our [medical] team members actually read a synthesis note. Because they can always come and ask us too, and have conversations.”
	“I think here, in the ICU at least, we are all together all day long…you are just in each other’s faces all day, so we have a lot of conversations that take place…the conversations might not always be documented, but I feel comfortable with how things get done. At least I feel like, if I have a concern ora question, it’s heard.”

Theme 2: Lack of Standardized Decision-Making Processes

This theme identified several processes related to the management of CVCs (see Table 1). The overarching theme was a perceived lack of a standardized process, or a lack of an awareness of agreed-upon indications for CVC necessity, despite recent incorporation of these standards into a safety checklist. At the time of the interviews, the safety checklist was a new tool, and thus, the process was not yet formalized; however, we also identified existing unit standards that were not well understood.

Two salient findings were that participants provided inconsistent information on routines related to CVC care, stating variations in suggested timeframes for dressing changes, tubing changes, and infusion hub replacements. It appeared that nurses referenced their own memory for these standards, rather than relying on a prompt in their documentation or by reading the institutional protocol. The second was that in the absence of a CVC, participants remarked that there was more deliberation regarding the use of antibiotics, presumably due to lack of access. Responses also reflected the perception that nurses and physicians have different perspectives on what necessitates a CVC, thus affecting the establishment of a set of standard criteria for team decisions.

Theme 3: Oral Communication Holds It Together

This theme reflected the idea that oral communication, as opposed to viewing data captured in the record, was the primary means for sharing information among the team and for making collaborative decisions (see Table 1). Oral communication is relied on and seen as beneficial when a patient condition changes or an urgent CVC management need arises. Checklist documentation served as a cognitive aid to prompt discussions during rounds that may not have occurred in the absence of an already identified and discussed need.

This idea that oral communication is the primary means for sharing information is also seen in participants’ statements regarding the input of the IV therapy nurse, noting that this was shared by “word of mouth.” Participants recognized that this “worked” in the ICU due to the medical team proximity and opportunity for conversations as needed. It was also noted that outside of the ICU, reliance on oral communication could prove problematic because communicating with the medical team involved paging physicians and waiting for them to call back.

DISCUSSION

This study aimed to describe what information is used in the management of CVCs and how this information is communicated among the healthcare team. As such, what information nurses deemed relevant to CVC management, where this information is found in the EHR, and how this information is shared with the interdisciplinary team were explored. Our final themes highlight opportunities to improve the implementation of CLABSI prevention strategies. First is ensuring that the daily review of catheter necessity is supported by relevant data from observations, assessments, and interventions captured in the record. Second is improving adherence to documented best practices of postinsertion catheter care through data-driven reminders in the record.^34–36 In addition, we identified a lack of consensus on indications for inserting and maintaining CVC access.

Daily Review of Device Necessity

Daily review of device necessity and removal of unnecessary catheters has been identified as one of five standard practices to prevent central line infections.^34,37 Previous research examining the implementation of this practice has demonstrated reductions in CLABSI rates in ICUs.³⁸ While research supports the use of daily rounds to evaluate device necessity, it is imperative that clinicians are able to access all relevant information, including historical information, from prior clinical shifts to make informed decisions. Due to the disparate nature of information captured in the record related to CVC management noted in this study, there may be challenges in referencing this information as a whole when deciding to continue or discontinue a CVC, and it would therefore be dependent on clinician recall and knowledge. Furthermore, even when nurses are present on rounds to provide information relevant to decisions about CVC access and removal, this information is not reliably captured in the record to inform future decisions and maintain continuity of care for the patient. This challenge is exacerbated in non-ICU settings where nurses may not be able to participate in interprofessional round due to higher patient ratios.³⁹

Ideally, an EHR would assist in developing a shared mental model among clinicians, even if from different professional groups such as nurses and physicians, by pulling together relevant information in a centralized location. Prior research regarding daily review of the necessity of CVCs has been focused in the ICU and has not fully addressed challenges related to non-ICU settings where CLABSI rates persist.^28,40 Our findings revealed reliance on oral exchanges as the primary means for sharing information. Consistent with previous research,⁴¹ nurses in our study believed that face-to-face communication worked in the ICU; however, they remarked that this would not likely be the case outside the ICU as daily face-to-face team communication is not typical in non-ICU settings. Others corroborate this perception,³⁹ further highlighting the need for standardized documentation practices surrounding CVCs to overcome the potential inconsistent oral communication outside the ICU. This further validates the need for EHRs designed to support information sharing across professional groups and to assist in developing displays that promote a shared mental model among the team. As such, we recommend the following functional requirements to be included in an interprofessional EHR module for CVC management (see Table 2).

Table 2.

Functional Requirements for the EHR Module

Prompt for daily review of device necessity

Reason to continue or discontinue documented by MD Reason to continue or discontinue documented by RN

Data presented to clinician to support CVC management decision (generated from EHR documentation)

Calculated number of days CVC in place based on date CVC inserted Last assessment by IV therapy nurse/RN discussing viabilityfor placement of a PIV or number of attempts to place PIV Last documented site assessment Medications requiring central access or total number of IV medications

Abbreviation: PIV, peripheral IV.

Postinsertion Central Venous Catheter Care and Maintenance

A review of the literature on CLABSI prevention efforts reveals a growing recognition that not enough emphasis has been placed on CVC care and maintenance after insertion to prevent CLABSI.^42,43 These concerns are bolstered by studies demonstrating increased infection rates in catheters with prolonged dwell times.^18,20,21 In addition, evidence demonstrates variations and inadequate adherence to best practices for catheter care.⁴¹ Our findings demonstrated varied practices of CVC maintenance including irregularities in nurses’ reports of catheter care practices (ie, frequency of dressing changes, tubing changes, and infusion hub [clave connector] changes). We recommend the use of clinical decision support tools within the EHR that support CVC decisions and that drive adherence to best practices related to the frequency of dressing changes, tubing changes, and other CVC care.⁴⁴

Standardized Criteria for Central Venous Catheter Access

While the newly implemented safety checklist included standardized prompts regarding CVC decisions, it had not yet been embraced as a truly interdisciplinary process, as indicated by nurses’ responses suggesting that this checklist “belonged to the physicians.” The nurses’ perception of ownership of the safety checklist also highlights the importance of sociotechnical theory as it relates to the implementation of information technology solutions. In order for new tools to achieve optimal impact, both the technical and social systems and subsystems need to be addressed.⁴⁵ Furthermore, additional research on an agreed-upon set of indications is needed.

CONCLUSION

While there are limitations due to the small sample size and single site, our findings provide a foundation for the development of EHR functional requirements that better support the management of CVCs and the prompt removal of unnecessary lines. Further research that includes both nurses and physicians, as well as clinicians who practice in non-ICU settings, will help to further validate and define what information is essential to making decisions regarding the continued need for CVC access. Effective interdisciplinary communication regarding CVC management must leverage relevant data captured in EHRs to support decision making and best practices that contribute to the elimination of CLABSIs.