Discussion
At the beginning of our study, SCD compliance was far less than expected. Data collection alone doubled compliance, with the improvement possibly attributable to the Hawthorne effect, whereby behaviour is modified by the awareness of being observed. Implementation of the best practice nursing alert resulted in further improvement in SCD compliance that was sustained, despite breaks in surveillance. SCD compliance nonetheless remained just below 50%. Implementation of patient-directed infographic to enhance patient engagement ultimately improved SCD compliance to over 50%. At best, this study achieved 55% compliance with prescribed SCDs. While it may be difficult to conceptualise anything less than full compliance, SCD compliance levels in our study are on par with other studies where compliance was reported to be 58% in a postoperative, gynaecological population19 and 53% in trauma patients at another tertiary centre.20
The various stages of this project highlighted how multifaceted the barriers to compliance are, and thus the need for the neurosurgeons to lead a multifaceted approach to interventions not only addressing equipment issues but also nursing and patient engagement. Where failure to deliver tubing with the SCD machine was rectified early on in this study, delays in delivery of the intact apparatus were especially felt by nursing staff. The machine used at our institution, the Kendall SCD 700 series compression therapy pump, is customisable, allowing for calf versus foot compression, setting total duration and unilateral settings. This degree of customisation increased the complexity and frequency of alarming in the setting of malfunction. This led to an increased tendency for the machine to simply be turned off by staff to make it stop alarming. Solutions for this problem could include greater education beyond uses of the power button or use of a non-customisable machine, such as the ALP—alternating leg pressure machine where the only options are power on and power off.
Distribution of the best practice alert achieved significant and sustained improvement highlighting the impact of nursing engagement and collaborative implementation of measures to improve SCD compliance. Despite a promising boost, it was met with less than hoped for returns, possibly due to nursing turnover and staff heterogeneity, with the nursing staff consisting of a mix of veterans, travellers and floats from other units. Staff heterogeneity required repeat redistribution of best practice alerts, regular physician-to-nurse communication, patient engagement and nurse to nurse focused communication concerning use of the SCD machine (figure 2).
Figure 2RN-prepared note at shift change. RN, registered nurse.
After nurse-focused interventions achieved their maximal effect, patient-focused interventions that allowed for achieving maximal SCD compliance rates were implemented. Creation of an SCD infographic to be given to patients likely acted in multiple ways to enhance patient engagement. Distribution of the infographic at the time of surgical scheduling for elective cases allowed patients to become familiar with SCDs ahead of hospitalisation. Distribution of the infographic to patients while on stepdown thus prompted attendings, residents and physician assistants to reinforce the importance of SCDs with patients, such that it became more apparent to patients that this constituted part of the ‘doctor’s orders’. Additionally, the process of sharing the infographic with patients also served as another reminder to nursing staff, of the need and/or importance of SCD use. Finally, this intervention provided patients with a way to take some control over a portion of their care.
Strengths of this study included its prospective design and clear, easy-to-implement, reproducible interventions as provided. The project took place at a tertiary medical centre with a patient population characteristic of large urban teaching hospitals so results are likely generalisable to other similar populations. The length of observation spanned multiple months with breaks in observation to allow discernment between a Hawthorne effect21 and lasting systemic impact. There may have been an initial change in nursing practice after daily study observation, highlighting the potential for physician providers to inspire system change. In this case, it would be expected that compliance would fall back to below the initial 20% after a break in surveillance. However, an improvement up to 40% was sustained, suggesting that lasting habits were made. Where other studies have failed to engage nursing staff early,22 much of this study’s success can be attributed to working in collaboration with nursing to identify barriers, devising and implementing the interventions. It was felt that early nursing buy-in was crucial to the improvement achieved since nurses are responsible for executing the orders placed and more likely to do so when engaged as part of the provider team. It was also felt that the creation of an SCD-specific best practice nursing alert and its incorporation as a routine aspect of nursing and CNA education would enhance long-term sustainability of compliance beyond its original introduction. The interventions in this study also included patient and physician engagement measures that not only enhanced compliance during the study but also raised awareness of the need for continued surveillance. By taking a multipronged and/or multidisciplinary approach and incorporating interventions at various levels to improving compliance, a system of checks and balances was created in order to enhance future sustainability.
This study, however, is not without its limitations. The duration for which SCDs should be used remains unclear. Where the American College of Obstetricians and Gynecologists recommends that devices be ‘used continuously until ambulation and discontinued only at the time of hospital discharge’, the practice bulletin also directs that devices ‘be continued until the patient is fully ambulatory’ a concept that is vague and does not give guidance as to how much time fully functioning patients need SCDs in place when not ambulating.23 Studies have reported that SCDs should be functioning for 18–21 hours a day to achieve optimal effect.24 This goal disregards how much time patients may be out of bed, ambulating to a chair, the restroom, working with physical therapy or undergoing bathing or linen changes.
Because our study included only once daily observations, we cannot speak to the total amount of time patients spent with SCDs in place. Some studies have used timers to pinpoint total duration of use25 or made six observations a day,20 neither of such practices was feasible in this study. However, it should be noted that the timing of our observations, between 6:00 and 7:00 hours, was best able to reflect overnight use when patients were most expected to be in bed and not up and about, getting changed, working with physical therapy or other activities. Another limitation of this study is that there are no data on whether the improved SCD compliance achieved better DVT prophylaxis and fewer VTE events, noting the degree of mismatch, where the incidence of an asymptomatic DVT detected by duplex ultrasound is 0.9% vs 2.6% of patients developing symptomatic PE.26 Without routine ultrasound and CT pulmonary angiograms, true incidence and the impact of improved SCD compliance cannot be fully appreciated. However, more than half of blood clots occurring after discharge have been directly linked to a recent hospitalisation or surgery.27 Furthermore, this study indicated that 40% of postoperative patients in the Department of Veterans Affairs hospitals have suffered VTEs while inpatient. Consequently, the need to mitigate inpatient risk cannot be overstated, as the full clinical significance of improved compliance may go unrecognised.
Lastly, this study aimed at improving compliance by changing the practices of staff and patients who were functional enough to engage in their own healthcare, in addition to addressing operational issues and increasing physician awareness. The study did not examine changes in knowledge-base stemming from the educational aspect of the implemented interventions. For example, with respect to nursing, it remains unclear if the best practice alert best addressed causative knowledge gaps, or if its distribution only reinforced the need for order compliance. Further study could examine the impact of the best practice alert on nursing perceptions of the SCDs and/or additional barriers to nursing compliance, including the use of additional educational tools.
Additional work is needed to further improve SCD compliance that ultimately remains low. This could be directed towards improvement in any of the areas impacting compliance and will likely require a higher level of system-wide interventions and further team-led changes, with physicians needing to maintain the highest levels of engagement. For example, a different SCD machine could be used with simpler display, less frequent alarms and/or wireless sleeves without tubing. The best practice nursing alert could be discussed by physicians and reviewed at hospital-wide nursing conferences, rather than those limited to the stepdown floor, increasing global nursing awareness of SCD compliance. At another institution, the authors are studying incorporating the best practice alert into the electronic medical record and making SCDs an automated order for postoperative patients. Finally, the patient-directed infographic could also be included in the hospital’s standardised preoperative pamphlet for elective cases and could also be posted along with the standard posters reminding patients about incentive spirometry and/or the iCOUGH protocol.
