Booster session prescription and outcomes in adults with chronic low back pain: Secondary analysis of a randomized clinical trial

Abstract

Background:

Booster sessions are suggested to maintain self-management behaviors and treatment effects in chronic low back pain (LBP) interventions, but the effects of boosters on outcomes and the best parameters for booster prescription are unclear.

Objectives:

(1) Compare booster prescription for two LBP treatments in an RCT where prescription was based on self-management program independence, (2) Determine if participant-specific variables predict requiring additional boosters, (3) Explore effects of boosters on outcomes in those requiring additional boosters.

Methods:

Secondary analysis of an RCT where 154 participants with LBP were randomized to motor skill training (MST), MST+Boosters (MST+B), strength/flexibility exercise (SFE), or SFE+B. This analysis focuses only on the +Boosters groups (age: 40.1±11.2 years, 49 female, LBP duration 9.8±8.8 years). Participants received MST or SFE and six months later received up to 3 boosters. Self-management program independence was assessed at the first booster, and those who were not independent required additional (>1) boosters. Chi-square tests were used to analyze booster prescription. Logistic regression analyses were used to examine predictors of requiring additional boosters. Descriptive statistics were calculated for outcomes for participants who did and did not require additional boosters.

Results:

MST+B and SFE+B did not significantly differ in returning for the first booster, χ²(1)=1.76, p=0.185. SFE+B were over 10 times more likely to require additional boosters than MST+B; OR = 10.9, 95% CI=[3.1, 38.1]. No participant-specific factors were statistically related to needing additional boosters. Attending additional boosters did not appear to change pain or function.

Conclusions:

Intervention type, not participant-specific factors, predicted the need for additional boosters. Attending additional boosters did not appear to change pain or function in the current sample.

Introduction

Chronic low back pain (LBP) is a prevalent, costly condition and is the leading cause of disability in the world.¹ Many rehabilitative interventions for LBP have been examined with positive effects on pain and function, including exercise, manual, and behavioral therapy.^2–5 Unfortunately, the largest effects are typically attained at short-term follow up.^4,6,7 Given that LBP is a condition with a course of persistent symptoms and disability for many, a goal is to prolong the positive effects of interventions over time, and self-management programs of exercise, education, or psychological interventions are suggested following an initial intervention.⁸ Because long-term adherence to these programs is challenging, boosters, or additional sessions provided periodically following an initial intervention, have been suggested as a method for maintaining self-management behaviors and treatment effects in people with LBP.^8–10

Some studies have examined boosters for people with chronic LBP.^11–19 However, in most studies, boosters were provided without a “no booster” comparison group, making it impossible to determine if there was a booster effect.^14–19 The few studies that compared identical interventions with and without boosters did not identify a booster effect on pain and function.^11–13 Additionally, the most appropriate number of booster sessions is not clear. In previous studies, there was substantial variation in the number of boosters ranging from 1 to 24 sessions and the same number of sessions was prescribed for all participants, regardless of status at the time of the booster.^11–19 Since providing additional treatment is costly, it would be ideal to deliver the fewest sessions possible at a relevant time point to maintain self-management behaviors and treatment effects, which might vary per individual. If we could predict who would need additional boosters, then we could focus resources on those individuals.

This study is a secondary analysis of data from a randomized clinical trial (RCT) comparing motor skill training (MST), MST+Boosters (MST+B), strength and flexibility exercise (SFE), or SFE+B for chronic LBP.²⁰ In the primary analysis, there were no statistically significant effects of boosters on pain or function (i.e., +B groups compared to other groups).²⁰ However, participants received different numbers of boosters based on their ability to perform their self-management program independently. Those who were not independent at the first booster were prescribed additional booster sessions (>1 booster up to 3 maximum; FIGURE 1). This design is novel in the study of boosters because it includes a no booster comparison group and individualizes the number of boosters based on program independence. The purposes of this exploratory study, therefore, are to (1) compare booster prescription in the MST+B and SFE+B groups, (2) determine if there are variables that predict who needs additional boosters, and (3) explore the effects of boosters on pain and function in people who needed additional boosters.

FIGURE 1.

Flow chart illustrating the prescription of boosters at booster session 1. Participants who were independent with their self-management program at the first booster were prescribed no additional boosters (1 booster), and participants who were not independent were prescribed additional boosters (>1 booster).

Methods

Design

This is a secondary analysis of a single-blind, RCT where participants with chronic, non-specific LBP were randomized to MST, MST+B, SFE, or SFE+B.²⁰ Outcome assessors were blinded to treatment assignment, but treating therapists were not. All testing and treatment occurred at Washington University. The study was approved by the Washington University Institutional Review Board (IRB ID# 201205051) and all participants provided informed consent. Individuals with chronic LBP (n=154) were recruited from the Saint Louis, Missouri, United States metropolitan area from December 2013 to August 2016. The sample size was determined for the primary aims of the RCT by an a priori power analysis to detect a difference of 6 points on the modified Oswestry Disability Questionnaire (MODQ)²¹ with 80% power.²⁰

Participants

154 participants were enrolled in the trial. Participants were included in the trial if they were 18–60 years old with chronic, non-specific LBP for at least 12 months with a moderate level of disability and were excluded if they had a BMI greater than 30, any specific cause of LBP, were pregnant, or involved in LBP-related worker’s compensation, disability or litigation. Full inclusion and exclusion criteria are in TABLE 1. The sample for this secondary analysis included the 76 participants who were randomized to boosters.

TABLE 1.

Inclusion and exclusion criteria for the trial.

Inclusion Criteria 18–60 years old Chronic low back pain (LBP) for ≥ 12 months Current LBP and not in an acute flare-up Modified Oswestry Disability Questionnaire ≥ 20% ≥ 3 LBP-limited functional activities Able to stand and walk without assistance Able to comprehend and read English and sign a consent form.

Exclusion Criteria BMI > 30 Neurologic disease requiring hospitalization Active cancer LBP etiology other than the lumbar spine Pregnancy LBP-related worker’s compensation, disability or litigation Fibromyalgia Marfan syndrome Graves’ disease Inability to classify LBP Any specific cause of LBP History of spinal fracture or surgery Frank neurological loss Pain or paresthesia below the knee.

Intervention

In the initial treatment phase, participants received 6 weekly, 1-hour sessions of MST or SFE. MST was person-specific training using motor learning principles to change movement and alignment patterns during LBP-limited functional activities.^20,22,23 SFE was exercises following ACSM guidelines to improve trunk muscle strength and trunk and lower limb flexibility in all planes. Full descriptions can be found in the primary outcome manuscript, Supplement 2.²⁰ At each visit, participants were assigned a program of their prescribed treatment to perform between sessions. At each subsequent visit, a reliable performance-based measure (inter-rater reliability, performance: % = 91 and k_w = 0.72 [95% CI 0.47–0.97]) was used to assess if participants were independent in their program and could be progressed.²⁴ For this measure, independence was defined as the ability to (1) verbalize the key concept of the exercise/skill correctly without cues, (2) perform the exercise/skill correctly without cues for (3) the prescribed number of repetitions (FIGURE 2). If participants met all 3 criteria, they were considered independent; otherwise, they were not. At the final treatment phase visit, participants were prescribed a self-management program of exercises/skills in which they were independent and encouraged to continue the program regularly for long-term management of their LBP. Six months later, participants attended a laboratory visit where they were informed of their booster assignment (booster/no booster). The 6-month time point was chosen a priori because that is when participants experienced a worsening of outcomes in a previous trial of an exercise and functional activity training intervention for people with chronic low back pain.²⁵ This secondary analysis is focused only on those participants randomized to boosters to examine characteristics of those who may need additional boosters and the effects of additional boosters on function, pain, and adherence.

FIGURE 2.

Criteria for independence in each exercise/skill in the self-management program with examples for participants in the strength and flexibility exercise (SFE) group and the motor skill training (MST) group. If participants met all 3 criteria for each exercise/skill they were prescribed, they were considered independent; otherwise, they were not.

At the first booster, each participant’s independence in their self-management program was assessed using the same performance-based measure used during the treatment phase (FIGURE 2). If the participant was independent, no additional boosters were prescribed. If the participant was not independent, then up to 2 additional sessions were prescribed weekly until the participant was independent or reached the 3-session maximum (FIGURE 1). We chose program independence as the criterion for determining who required additional boosters because a primary aim of boosters is to help promote long-term behavior change and self-management.^8,26–28 The 3-session maximum was set to minimize costs while providing enough treatment to boost self-management program independence. Treating therapists assessed the participants’ independence in their program at the first booster visit and were, therefore, not blinded to treatment assignment. However, the therapists that treated the MST group were not the same therapists as those that treated the SFE group.

Measures

The primary outcome for the RCT was the MODQ, a validated measure of LBP-related functional limitation with higher scores indicating worse function.²¹ Secondary outcomes were the Numeric Pain Rating Scale (NRS; 0–10) for average and worst LBP in prior 7 days,²⁹ adherence to the home program (participant-reported weekly adherence, 0–100%, during the treatment and booster phase and monthly adherence, 0–100%, during the follow-up phase),^25,30 LBP flare-ups,^31,32 SF-36 Physical and Mental Component Summary scores,^33–35 absenteeism and presenteeism,^36,37 LBP-related treatment, medication, and equipment use, fear-avoidance beliefs^38,39 and satisfaction with care.⁴⁰ All measures (except adherence to treatment and satisfaction with care) were collected at baseline. At each treatment session or booster session, MODQ, pain, and adherence to treatment (for visits 2–6 or boosters 2–3) were collected. After the initial treatment phase participants completed a subset of self-report surveys monthly for 12 months. All data were collected using REDCap.^41,42 Outcome assessors were blinded to treatment assignment, but treating therapists were not. For this secondary analysis, additional measures included attendance at the first booster session and the number of boosters prescribed for those who attended the first booster.

Statistical Analysis

Booster attendance and prescription

All statistical analyses were conducted in R v4.1.2.^43,44 To analyze attendance at the first booster session (yes/no) and the number of boosters as a function of Group (MST+B versus SFE+B), we used a Chi-square test with Yates’ continuity correction, which reduces over estimation in small samples.⁴⁵

Moderators of booster prescription

Next, we used logistic regression to model the probability of needing additional boosters for those participants who attended the first booster session (i.e., participants who did not attend any boosters were excluded from these analyses). Given our limited sample size, we started with a model testing the effect of booster prescription (additional boosters versus no additional boosters) regressed onto Group (MST+B versus SFE+B) as a base model. We then tested a series of individual models adding the effects of age, sex, duration of LBP, adherence, MODQ, and average pain as predictors controlling for group.

Exploring differences in outcomes in those who needed additional boosters

Finally, we were interested in how booster attendance might affect adherence, MODQ, and pain from the booster phase to the post-booster phase (the six months following the last booster) of the study. However, given the lack of statistical power and the small, uneven group sizes, we decided not to calculate inferential statistics for these changes. Instead, we present descriptive statistics for adherence, MODQ, and pain for participants who needed additional boosters versus no additional boosters in each group (MST+B versus SFE+B) during the booster phase and the post-booster phase.

Role of the Funding Source

The funders played no role in the design, conduct, or reporting of this study.

Results

Participant characteristics at baseline at the beginning of the trial

The sample for these analyses included the 76 participants (mean ± SD age: 40.1±11.2 years, 49 female) who were randomized to boosters. At the beginning of the trial the participants had an average MODQ of 30.7±9.5, average pain rating in the past week of 4.7±1.7, and LBP duration of 9.8±8.8 years (TABLE 2). Forty participants were randomized to MST+B and 36 were randomized to SFE+B. The characteristics of participants in the MST+B group and SFE+B group were not significantly different with the exception of sex and worst pain in the prior 7 days. There were significantly more females in the MST+B group (p=0.017) and the worst pain was significantly higher in the SFE+B group (p=0.044).

TABLE 2.

Baseline characteristics for individuals randomized to booster treatment

Characteristic	Motor Skill Training (n=40)	Strength & Flexibility Exercise (n=36)
Female – no. (%)	31 (77.5)	18 (50.0)
Age (y)	39.2 ± 11.0	41.1 ± 11.4
Duration of LBP (y)	9.3 ± 9.2	10.4 ± 8.5
Numeric Pain Rating Scale†
Current	4.0 ± 1.9	4.0 ± 1.6
Average (7 day)	4.7 ± 1.4	4.7 ± 1.9
Worst (7 day)	6.1 ± 2.1	6.9 ± 1.3
Modified Oswestry Disability Questionnaire‡	31.5 ± 11.1	29.4 ± 7.5
Current LBP Medication Use – no. (%)§	27 (67.5)	23 (63.9)
Equipment Use – no. (%)¶	38 (95.0)	28 (77.8)
Fear-Avoidance Beliefs Questionnaire – physical activity subscale score#	14.4 ± 6.2	14.2 + 4.4
Fear-Avoidance Beliefs Questionnaire – work subscale score#	11.3 ± 9.1	13.5 ± 8.3
Absenteeism (days)††	3.0 ± 5.2	2.9 ± 5.0
Stanford Presenteeism Scale‡‡
Work Output Score (%)	88.9 ± 13.4	83.1 ± 18.7
Work Impairment Score	19.4 ± 5.5	21.2 ± 6.4
SF-36 Physical Component Score§§	42.8 ± 7.0	41.3 ± 6.3
SF-36 Mental Component Score§§	50.6 ± 11.6	50.9 ± 9.4

NOTE. Baseline is at the beginning of the trial. Values are mean ± SD or no. (%).

Abbreviations: y, years; no., number; LBP, low back pain;

^†Numeric Pain Rating Scale ranges from 0–10 with lower scores indicating less pain.

^‡Modified Oswestry Disability Questionnaire Scores range from 0–100% with lower scores indicating less disability.

^§Number of people who report medication use for LBP. Includes non-prescription and prescription medication.

^¶Number of people who report equipment use for LBP. Includes orthotics, braces, mobility aids, traction devices, massagers, modalities, furniture, bedding, bathroom aids, or footwear.

^#Fear-Avoidance Beliefs Questionnaire physical activity subscale scores range from 0–24 and work subscale scores range from 0–42 with lower scores indicating lower fear avoidance.

^††Number of days in the past 4 weeks a person was kept from his usual activities due to LBP. Scores range from 0–28 days.

^‡‡Stanford Presenteeism Scale Work Output Score is the participant’s estimate of the percentage of his usual productivity level during work over the past 4 weeks (0–100%). Work Impairment Score ranges from 10–50 with 10 indicating the lowest degree of impairment.

^§§36-Item Short Form Health Survey (SF-36) Physical and Mental Component Summary Scores range from 0–100 with higher scores indicating better physical or mental health.

Booster prescription

As shown in FIGURE 3, of the 40 participants in the MST+B group, 33 returned for the first booster and 7 did not (5 discontinued trial participation prior to booster assignment, 1 refused to attend the 6-month lab visit but continued trial participation, and 1 refused boosters but continued trial participation). Of the 36 participants in the SFE+B group, 24 returned for the first booster and 12 did not (5 discontinued trial participation prior to booster assignment, 1 refused to attend the 6-month lab visit but continued trial participation, and 6 refused boosters but continued trial participation). The difference in returning for the first booster between groups was not statistically significant, χ²(1)=1.76, p=0.185. Among participants who returned for the first booster, 27/33 (81.8%) MST+B required no additional boosters (i.e., were independent with their program at the first session) compared to 7/24 (29.2%) SFE+B, a statistically significant difference χ²(1)=13.89, p<0.001. All of the participants in both groups were independent in self-management program performance in ≤ 3 booster visits except 1 SFE participant.

FIGURE 3.

Number of booster sessions as a function of group. MST = motor skill training; SFE = strength and flexibility exercise. Participants who attended 0 booster sessions includes all participants randomized to boosters who did not attend the first booster session (randomized to boosters but discontinued trial participation prior to booster assignment, refused to attend the 6-month lab visit where they were informed of booster assignment, or refused boosters). Participants who required no additional boosters were independent in their self-management program at the first booster and participants who required additional boosters were not.

Moderators of booster prescription

Logistic regression comparing who required additional boosters versus no additional boosters again reflected that SFE+B participants were over 10 times more likely to need to return for multiple sessions compared to MST+B participants, OR=10.9, p<0.001, 95% CI = [3.1, 38.1]. Controlling for group, we added other variables in a series of individual models to see if they were related to requiring additional boosters. Neither age (OR=1.04, p=0.173, 95% CI = [0.98, 1.11]), identifying as male (OR=2.63, p=0.174, 95% CI = [0.66, 10.36]), the duration of LBP (OR=1.03, p=0.450, 95%CI = [0.96, 1.01]), nor adherence to the self-management program prior to the booster (OR=0.99, p=0.386, 95%CI=[0.96, 1.01]) were significantly related to needing additional boosters. For this analysis, we also were able to look at how MODQ and pain at the first booster related to needing additional boosters. However, neither MODQ (OR=1.02, p=0.490, 95%CI=[0.96, 1.09]) nor pain (OR=1.23, p=0.388, 95%CI = [0.77, 1.95]) were significantly related to needing additional boosters.

Exploring differences in outcomes in those who needed additional boosters

Descriptive statistics for adherence, MODQ, and pain for the booster phase and the post-booster phase (6 months following the booster) for those who required additional (>1) boosters and no additional (1) boosters are presented in TABLE 3. Mean differences and layered confidence intervals (90 and 97%) for groups who required additional (>1) boosters versus no additional (1) boosters in the post-booster phase for adherence, MODQ, and pain are presented in FIGURE 4.

TABLE 3.

Descriptive statistics for adherence, modified Oswestry Disability Questionnaire (MODQ), and pain across the different phases of the experiment

	Measurement
	Adherence†		MODQ‡		Average Pain (NRS)§
Group	Booster	Post-Booster	Booster	Post-Booster	Booster	Post-Booster
MST
No additional boosters (1 Booster), n=27	73.1 (22.5)	77.0 (17.6)	9.4 (9.0)	7.87 (7.4)	1.19 (1.33)	1.28 (1.14)
Additional boosters (>1 Booster), n=6	75.2 (14.1)	84.2 (9.8)	14.0 (8.7)	17.4 (17.9)	1.67 (1.03)	2.07 (1.62)
SFE
No additional boosters (1 Booster), n=7	46.9 (34.3)	57.2 (32.1)	20 (8.87)	19.8 (9.0)	2.14 (1.21)	2.04 (1.20)
Additional boosters (>1 Booster), n=17	31.4 (23.2)	50.0 (17.8)	19.5 (10.5)	20.0 (13.2)	2.41 (1.58)	2.30 (1.63)

Note that all values are presented as Mean (SD). For adherence, the “Booster” phase refers to adherence in the month immediately before the first booster session. For MODQ and average pain symptoms the “Booster” phase averages across all boosters attended, and then averages across participants. In the Post-Booster phase (the six months following the last booster), we averaged across all available data points within a subject, and then averaged across participants to get a group level average over time. MST = motor skill training; SFE = strength and flexibility exercise.

^†Adherence scores range from 0–100% with higher scores indicating better adherence.

^‡MODQ ranges from 0–100% with lower scores indicating less disability.

^§Numeric Pain Rating Scale (NRS) ranges from 0–10 with lower scores indicating less pain

FIGURE 4.

Mean differences and layered confidence intervals (90 and 97%) for adherence, Modified Oswestry Disability Questionnaire scores (MODQ), and pain (NRS) between groups who required additional (>1) versus no additional (1) boosters in the Post-Booster phase (the six months following the last booster). The darker bars indicate the 90% confidence intervals, and the lighter bars indicate the 97% confidence intervals. MST = motor skill training; SFE = strength and flexibility exercise. Adherence scores range from 0–100% with higher scores indicating better adherence. MODQ scores range from 0–100% with lower scores indicating less disability. Numeric Pain Rating Scale (NRS) values range from 0–10 with lower scores indicating less pain.

Discussion

Our first purpose was to compare booster prescription in the MST+B and SFE+B groups. There was not a significant difference in attending the first booster by group. However, there was a difference between groups in their independence at that session and thus in the number of boosters prescribed. SFE+B participants were over ten times more likely to require additional boosters because they were not independent in their self-management program compared to MST+B participants. Current clinical practice guidelines emphasize the importance of self-management in chronic low back pain.⁴⁶ Our findings suggest that MST results in better performance of a program 6 months following treatment than SFE. This difference may be because MST was designed specifically to facilitate learning. This is important because improved self-management of a chronic condition could enhance health care resource efficiency. Additionally, in our primary analysis of the trial, we previously reported that participants who received MST showed greater short-term and long-term improvements in function than those who received SFE.²⁰

Our second purpose was to determine if there are variables that predict who will need additional boosters. When controlling for group, none of the participant-specific variables of age, sex, or duration of LBP were significantly related to needing additional boosters. Additionally, pain, function, and adherence were not significantly related to needing additional boosters. Since we did not find a relationship between needing additional boosters and LBP-related outcomes, one could question our criterion for determining who needs additional boosters. We used independence in the self-management program as the criterion because it was the primary method for managing low back pain (LBP). We did not use a change in pain or function to prescribe additional boosters, because the goal of the boosters was to help maintain self-management over the long-term rather than to wait for a worsening in pain or function. Our rationale is consistent with previous studies that describe the purpose of boosters as promoting long-term independence and behavior change.^{13,14,16–19} While independence in the program is one way to measure self-management behaviors, these behaviors are complex. Therefore, utilizing other metrics, such as patient adherence to the program, could provide a more comprehensive assessment of self-management behavior. Different from our study, in previous studies of boosters for LBP the same number of sessions was prescribed for all participants.^11–19 It is possible that a criterion other than program independence would better predict who would benefit from additional boosters, for example, adherence, pain, or function – this could be explored in future studies because, to our knowledge, no other study has used a criterion to determine how many boosters to prescribe.

We were not able to statistically test differences in outcomes for participants who needed additional boosters versus no additional boosters. Thus, the values presented are a description of the current sample, rather than generalizations to the larger population. In the SFE group, people who needed additional boosters appear similar to people who needed no additional boosters in terms of pain and function both during the booster and post-booster phase. It appears that for those who were not independent, additional treatment to “boost” their performance did not improve their pain or function. In contrast, in the MST group, people who needed no additional boosters appear to be doing better than those who needed additional boosters, and better than people in the SFE group who attended any number of boosters. We previously reported that compared to no boosters, there was no effect of boosters on pain or function when including everyone who was randomized to boosters.²⁰ Here we explore those that needed additional treatment per our criterion and there do not appear to be changes in pain or function in the 6 months following the booster. In this sample, booster sessions of the original treatment did not appear beneficial, even for participants who were no longer independent in performing their self-management program. Although these findings cannot be generalized to the broader population, they are clinically meaningful within this sample and suggest that alternative booster models warrant further exploration. It is possible that individuals may require different treatments 6 months post-treatment. Trials with innovative designs, such as Sequential Multiple-Assignment Randomized Trials, which provide additional treatment to non-responders during the initial treatment phase, are currently being explored in LBP. A similar approach could be explored for boosters.^47,48

In addition to considering who may need additional boosters and what type of treatment should be prescribed, when to provide boosters is also a consideration. We chose 6 months following the initial intervention because that is when participants in a previous trial declined in function.²⁵ However, participants in the current trial did not experience the same decline in function from 6–12 months.²⁰ Perhaps boosters would have a greater effect if the timing of the boosters was individualized, since the changes in function observed in both trials were group averages rather than individual trajectories. This is particularly important to explore further since most previous studies of boosters in LBP provide no specific rationale for the timing of sessions.^{11,12,14–17,19}

Limitations

We had a limited sample size to explore booster attendance and the lack of a significant difference in attending the first booster should not be interpreted robustly. For the moderators of booster prescription, the study was not powered to detect associations between additional boosters and prognostic variables, and these analyses are exploratory. Additionally, there were uneven and small numbers of participants in the groups who required additional boosters vs no additional boosters. Thus, we were unable to statistically explore differences between the groups. Since we specifically assessed MST and SFE, we cannot generalize our findings to other treatments for chronic LBP. However, SFE is one of the most recommended and commonly prescribed treatments for chronic LBP.⁴⁹ Additionally, our findings may not be generalizable to individuals with LBP who have BMI >30, specific LBP conditions, or those receiving LBP-related worker’s compensation. This is particularly relevant since higher body weight and physically demanding jobs have been identified as predictors of poorer outcomes in people with LBP.⁵⁰

Conclusions

There was not a significant difference in returning for the first booster between the MST and SFE groups, but MST participants were over 10 times more likely to be independent in their self-management program 6 months after treatment and thus needed fewer additional boosters than SFE participants. None of the additional factors we explored were significantly associated with program independence. If the goal of LBP rehabilitation is to achieve long-term independent performance of a self-management program, our findings suggest that clinicians may consider MST, a treatment designed to facilitate motor learning, over SFE. While we could not statistically test differences in outcomes between those who required 1 vs >1 booster, attending additional boosters did not appear to change pain or function in the current sample. These values are a description of the current sample and should not be generalized to the larger population. However, they do suggest that alternative booster models should be explored. Our study suggests that future studies of boosters should examine the criteria for determining who needs additional treatment, the timing of sessions, and what type of treatment should be provided.

Data Availability Statement:

Data available on request: The data presented in this study are available on request from the corresponding author.