Quality Improvement Approach to Increase Inpatient Pediatric Secondhand Smoke Exposure Screening

The project site was an 8-bed general pediatrics inpatient unit within an adult academic safety-net hospital and regional trauma center. A total of 62% of our pediatric patients are Latinx, and 10% of the patients are African American; in addition, 81% have Medicaid insurance, and 7% are uninsured. Although physicians on the unit were trained to screen patients for tobacco SHSe, this information was not documented in a standardized manner in patient charts. Nurses did not routinely screen for SHSe because the nursing intake form was focused on primary tobacco use. Chart review revealed that none of the nursing intake forms for patients admitted between June 2018 and August 2018 documented SHSe screening.

A team composed of medical students, nurses, and physicians who were particularly interested in quality improvement (QI) and smoking cessation worked together on this project. Additional stakeholders, such as unit clerks, nurse managers, and local leaders in smoking cessation efforts, were also engaged.

First, we sought to understand the root causes of the inadequate SHSe screening rates. In October 2018, we surveyed nurses, attending physicians, and residents who work on the inpatient pediatric floor to assess perceptions of and barriers to SHSe screening and intervention (Fig 1). Nurse-delivered screening has been shown to increase patient success in smoking cessation,^20,21  and the nurses embraced implementing the intervention. We found that, although 100% of nurses felt that SHSe screening was “Very Important” or “Important,” only 50% of them felt “Very Confident” or “Confident” in their ability to have conversations with patients and their families about the impact of smoke exposure on pediatric health, similar to that in previous studies.²²  The most frequently cited barriers to SHSe screening were the lack of a standardized screening tool, ambiguity about next steps when a patient screens positive for SHSe, and insufficient education about cessation resources.

With our intervention, we sought to address these barriers by doing the following:

1. Creating a SHSe screening tool: A paper-based screening tool was developed and acted as both a script and workflow checklist (Fig 2). The screening questions were based on recommendations from the CEASE framework.¹⁸  We incorporated this tool into the existing nursing intake form to ensure timely and systematic screening for each admission. At the time of the intervention, our hospital was changing electronic health records (EHRs). We piloted a paper-based system because the nursing admission workflow remained paper-based during this transition period. Additionally, the paper-based tool cost less and allowed for more adaptability in response to user feedback, compared with that of an electronic screener.

2. Designing a streamlined nursing workflow (Supplemental Fig 4): The unit clerk was responsible for attaching the screening tools to the nursing intake forms and collecting completed screening tools. After admission, the nurse used the screening tool to ask the caregiver about pediatric exposure to cigarette, e-cigarette, or marijuana smoke. If a patient screened positive for cigarette or e-cigarette SHSe, the caregiver was offered 3 smoking cessation resources. First, the nurse offered a referral to the California Smokers’ Helpline. We decided to have staff enroll all interested caregivers, rather than giving caregivers the link to register themselves because data suggest that proactive enrollment is associated with increased usage of the quitline.²³  Next, the nurse offered a “Quit Kit” created by our city’s Tobacco Free Project, which contained a variety of materials to help curb smoking cravings, such as a water bottle, mints, gum, lip balm, and educational materials. After completing the intake process, the nurse notified a physician for positive screen results. The physicians were all trained in CEASE, as well as additional motivational interviewing techniques, to provide smoking cessation counseling and an NRT prescription. If the patient screened positive for marijuana SHSe, the nurse notified a physician to offer education about the adverse health impacts of marijuana SHSe as well as harm-reduction techniques.

3. Administering nurse training sessions: At the start of project, we conducted 2 in-person training sessions on the new workflow, smoking cessation referral resources, and the impact of SHSe on children’s health by using slides (Supplemental Information) and role play activities.

In the weeks before implementation, nursing huddle boards were created by the QI nursing leads to track the screening data in real-time. The SHSe screening and smoking cessation referral protocol was implemented on April 15, 2019. Visual reminders to screen patients (saying “take a second to ask about secondhand smoke”) were placed in highly visible locations, including patient room doors, hand-washing stations, and nursing stations.

Our primary outcome measure was the monthly percentage of admissions screened for SHSe from May 2019 to April 2020. We cross-referenced completed SHSe screening forms with the daily general pediatrics nursing census to calculate screening rates. Our secondary outcome measure was the average annual percent of patients who screened positive for SHSe who were offered a referral to each smoking cessation resource (“Quit Kit,” California Smokers’ Helpline, physician cessation counseling). Process measures were the percent of all nurses who attended the workflow training and nurse confidence in SHSe screening over time, which was assessed by using Likert scale and free-response questions in 2 postimplementation surveys. As a balancing measure, we tracked nursing satisfaction and any negative impact to original workflow using postimplementation surveys. Another balancing measure was screening brevity (taking less than 3 minutes) to prevent nurses from reducing time spent on other patient care tasks.

We used run charts to track the rates of SHSe screening over time. Observed monthly screening rates were compared with our predetermined, arbitrary target of 70%. We omitted April 2019 from our analysis because the monthly screening rate for April included data before implementation. Descriptive statistics are presented as percentages and means. The percent exposed to each substance (eg, cigarettes) was calculated on the basis of the number of screening tools that had a response recorded for that particular substance. Patients for whom the screening tool did not have a response marked for a substance were excluded from the total patients screened, resulting in different denominators for each substance.

This project was exempt from review by the university’s institutional review board because it was a QI project and did not pose significant risk to patients.

During the first year of implementation, 97.2% of the 394 admissions to the pediatric unit were documented as being screened for SHSe. The documented monthly screening rates were ≥94% (Fig 3). Analysis of the admissions that did not have documented SHSe screening revealed that most of the patients were transfers from other units that do not screen for SHSe or were admitted for circumstances that made screening for SHSe challenging (eg, infant under the custody of Child Protective Services or patient admitted for injuries sustained from suspected child abuse).

Of the 383 admissions with documented SHSe screening, there were 364 unique patients and 19 readmissions. Our results revealed that 70 patients (19.2% of unique patients) had at least 1 exposure. Of the patients with at least 1 exposure, 15.7% were exposed to cigarettes or other tobacco products (n = 57), 5.6% to vaping or e-cigarettes (n = 20), 6.5% to marijuana (n = 22), and 34.3% (n = 24) had exposure to >1 substance (Table 1). Of note, of the 17.0% (n = 62) of unique patients who screened positive for SHSe to tobacco products and/or e-cigarettes, 5 patients had noncaregiver or family member exposures, such as friends at school or neighbors. Additionally, 5 of the patients who screened positive for SHSe were readmitted during the study period. A total of 4 patients were readmitted within the same month as the initial admission. One patient was readmitted a few months after the initial admission, and, after readmission, the child no longer screened positive for SHSe.

In terms of our secondary outcome measures, 45 caregivers were documented as being offered a “Quit Kit” and 37 accepted (representing 72.6% and 59.7% of positive screen results, respectively). Additionally, 33 caregivers (53.2% of positive screen results) were offered a referral to the California Smokers’ Helpline. A total of 16 referrals to the helpline were electronically entered for the caregivers of 14 patients, meaning that for 22.6% of patients who screened positive for SHSe, at least 1 caregiver agreed to be referred. The most commonly documented reasons that caregivers refused these resources were a lack of interest in smoking cessation and absence of the individual who smokes at the bedside. Nurses documented notifying a physician of a patient who screened positive for SHSe 62.9% (n = 39) of the time. The smoking cessation resource documentation rates are displayed for the intervention duration in Table 2.

In terms of our process measures, 100% of the unit nursing staff attended the workflow training before implementation. Our other process measure was nurse confidence in screening for SHSe, which we assessed by surveying nursing staff 1-month postimplementation (100% response rate) and 18-months postimplementation (100% response rate). A total of 100% of nurses at 1 month and 94% of nurses at 18 months reported that the screening tool “Increased” or “Significantly Increased” their confidence in screening patients and caregivers.

Our balancing measures were also assessed in the surveys. Nursing staff wrote that screening was a “quick process and families [were] willing to answer” the questions but that sometimes they would “need to go back and spend more time when [the] admission settled down.” Respondents also expressed satisfaction with the tool, writing, “Great idea! Always good to understand the social determinants of health affecting our patients.”

We successfully designed and implemented a comprehensive SHSe screening workflow, which increased our documented rate of screening from 0% preimplementation to 97.2% 1-year postimplementation. Monthly screening rates of at least 94% were throughout the intervention period, without any repeated training sessions after the implementation of the tool. These consistently high screening rates can improve patient outcomes by directly contributing to the denormalization of tobacco.¹⁹  Additionally, nursing staff maintained a positive attitude about our screening tool and felt that because of its brief nature this intervention did not negatively impact other patient care tasks, which were the key balancing measures of this project. Our sustained success highlights the importance of cohesively integrating new tools into existing workflows and collaborating with key stakeholders, including nursing leadership and administrative staff, at all stages of a QI project.

The screening forms provided important exposure data about our specific pediatric patient population: 19.2% of patients were exposed to cigarettes, e-cigarettes, or marijuana. The secondhand cigarette rate (15.7%) is higher than the average smoking rate in our city of 9.5%²⁴  but more aligns with the cigarette smoking rate of those with an annual household income of < $35 000 in the United States (21.4%).²⁵  A study in Colorado (where marijuana is also legal) revealed that 16% of children hospitalized for bronchiolitis had metabolites of marijuana smoke in their urine, which may suggest that our numbers are underreported.²⁶  Of all 3 of the substances screened for, vaping was reported the least. This may reflect that e-cigarette use rates decline with age and that those who use e-cigarettes are younger than our average caregiver.²⁷  Existing QI research has been focused on SHSe to tobacco products, so with the inclusion of e-cigarette and marijuana in our study, we provide unique exposure data, which are important both from a provider perspective as well as from a public health perspective, given the individual and systems level health impacts of SHSe.

Screening allowed nursing staff to identify patients and caregivers who would be most likely to benefit from multimodal smoking cessation resources, which was our secondary aim. With the data in Table 2, we illustrate that there is room for improvement both in terms of offering resources to 100% of eligible caregivers and increasing the percent of caregivers who accept the resources when offered. The most commonly documented reasons that caregivers refused these resources were (1) lack of interest in smoking cessation and (2) the individual who smokes was not present at the bedside during screening (eg, family member who was not present, neighbor, patient’s friends, and patient’s schoolmates). Relatedly, nurses sometimes did not offer smoking cessation resources if the individual who smoked was not present at the bedside. Another barrier to nurses offering smoking cessation resources was the need to prioritize a medically urgent aspect of the patient’s care. Often in these instances, the corresponding fields were left blank on the screening tool. If the nurses offered the resources later during the patient’s hospitalization but did not revise the screening tool, this would not be captured in the rates we have reported here. Referrals to the helpline were the resource with the lowest rate of being documented as offered by nursing staff, but our rates are still greater than of published data in which researchers report that pediatricians refer only 14.7% of caregivers who smoke to quitlines.²⁸  A referral to the helpline doubles an individual’s chance of quitting nicotine, so these 16 referrals are still a notable success, particularly because there were zero documented referrals made to the helpline in the year before implementation, and offer an opportunity for further improvement in the future.²⁹ 

In addition to the increase in “Quit Kit” and helpline referrals, the new workflow also facilitated interdisciplinary provider communication. Nurses documented that they notified a doctor for 62.9% of the patients who screened positive for SHSe, a statistic that could underestimate the actual notification rates if physicians were notified later in a shift after the screening paperwork had been filed or if nursing staff notified a physician but then forgot to document this step of the workflow.

There are a number of limitations of our project. First, it presents data from a single site: a safety-net hospital with a relatively low daily admission rate (average of ∼1 admission per day) that uses a paper-based nursing intake form. Second, our intervention is meant to initiate conversations around smoking and connect people who are interested in cessation to resources, but if the caregiver is not interested in quitting, this would not be an effective stand-alone intervention. Finally, it was beyond the scope of our project to individually follow up with the caregivers who received resources to determine smoking cessation rates, but we closely adhered to the CEASE principles that have previously revealed increased smoking cessation.³⁰ 

Future interventions can be aimed at optimizing smoking cessation resource distribution rates. New goals are to reach a 100% physician notification rate of a positive screen result and create a take-home helpline enrollment form that could be used if the caregiver who smokes is not in the hospital.³¹  Phone interviews with the caregivers who accepted resources could be used to assess cessation rates and engagement with additional resources, such as NRT. This future study may be used to provide data to address the currently inadequate rates of NRT prescription, as illustrated in a survey of pediatric hospitalists published in January 2021 that found that 72% of respondents never prescribe pharmacotherapy for smoking cessation.³²  It would be informative to evaluate the resource allocation rates to examine if there are any discrepancies on the basis of factors such as patient language, race and ethnicity, sex, or admission diagnosis. The California Smokers’ Helpline (which is rebranding as Kick It California in Fall 2021) recently expanded its marijuana-related offerings,²⁹  so this could be emphasized. Finally, although we chose to implement a paper-based method as a cost-effective way to establish proof of concept, this screening tool could feasibly be integrated into the EHR, which would enable adoption of this tool more broadly across institutions that use EHRs for nursing intake documentation.

We created and successfully implemented a paper-based SHSe screening tool that enabled the targeted delivery of smoking cessation resources in a way that could be replicated at other clinical sites with minimal resources, particularly at safety-net hospitals. Our pilot project will contribute to the existing QI research on this topic because of its inclusion of exposures beyond tobacco smoke, including e-cigarette and marijuana smoke, and its low-cost implementation within a safety-net hospital inpatient unit.

Thank you to Aisha Schexnayder, Pediatrics Unit Clerk, for creating and organizing a storage system for the completed patient screening forms.