Reducing Upper Extremity Precautions After Lung Transplant: The Clamshell Protocol Pilot Study

Abstract

Background and Purpose:

A less-restrictive clamshell precaution (CP) protocol for postsurgical upper extremity (UE) movement was implemented after lung transplantation (LTx) instead of traditional sternal precaution (SP) protocols. The purpose of this pilot study is to assess the safety of transitioning from the SP protocol to the CP protocol.

Methods:

This pilot study was a single-site, observational, cohort study at an academic medical center from March 1, 2021, to February 28, 2023. The CP protocol was implemented on March 1, 2022. Cohorts were determined by the protocol in place at the date of LTx. Safety is described by the frequency of adverse events (AEs), defined as sternal instability or chronic surgical wound infection within 6 months. Summary statistics were reported. Adverse event rates are reported with 95% Wilson score confidence intervals (CIs).

Results:

Demographics and clinical characteristics were similar between the cohorts. AEs occurred in 4 patients (10%), including 3 patients (18.8%, 95% CI: 6.6%–43.0%) in the SP cohort (n = 16) and 1 patient (4.2%, 95% CI: 0.7%–20.2%) in the CP cohort (n = 24). Those with AEs were older with a higher body mass index and comorbidity index score compared with those who did not have AEs.

Discussion:

In this pilot study of reducing UE movement restrictions after LTx, there was preliminary evidence of lower AEs, suggesting that reducing restrictions may be a safe alternative to the current standard of care.

Conclusions:

The results of this pilot study provide preliminary support for the safety CP protocol. However, because of the small sample sizes, further studies are needed to confirm this.

INTRODUCTION

Lung transplantation (LTx) is a life-saving procedure that improves pulmonary function and quality of life for those with end-stage lung disease.¹ Complications, including impaired surgical site healing, are a significant concern. The incidence of poor sternal healing is particularly high, reaching up to 36% following the clamshell incision, which is the most invasive surgical incision used for LTx.^2,3 These complications can lead to pain, functional limitations, diminished health-related quality of life (HRQOL), and the need for expensive and painful surgical revisions.^3-5

To minimize the risk of complications, surgeons prescribe postsurgical precautions primarily involving upper extremity (UE) movement restrictions.^6,7 The precautions used vary by institution but often replicate the sternal precautions used in cardiac surgeries that require a median sternotomy.^6,7 However, notable differences exist between the median sternotomy and the clamshell incision. Unlike the vertical sternotomy incision, most LTx techniques involve horizontal sternal transection, affecting different muscle groups and altering the movements that strain the surgical site. Consequently, traditional sternal precautions may not fully address the unique demands of LTx recovery. Recovery after LTx may be further affected by UE movement precautions as precautions greatly impact physical therapy (PT) and pulmonary rehabilitation (PR), which aim to improve strength and aerobic capacity without increasing the risk of healing complications.^6,8,9 To date, the priority has been to minimize the risk of healing complications, leading to the use of restrictive sternal precautions that restrict UE range of motion and prohibit any lifting >5 pounds (lbs) for the duration of the protocols.⁷ Despite the common use of sternal precautions for 6 to 10 weeks after surgery, there is little evidence supporting their use.¹⁰ No work has been done to establish the efficacy of sternal precautions following LTx. In addition, recent work found that less-restrictive protocols for median sternotomy improved function and HRQOL without increased rates of healing complications.^10-13 This lack of evidence specific to LTx and the emerging evidence that supports reduced restrictions in other patient groups necessitates that we reconsider the postsurgical precautions in place after LTx.

As a result of this lack of evidence, the PT staff at the study center used the existing research and expert opinion to develop a less-restrictive UE mobility protocol for use during PT and PR after LTx and implemented it as a quality-improvement initiative.^14-16 The purpose of this pilot study is to preliminarily assess the safety of the novel less-restrictive protocol compared with the traditional sternal precautions. The safety of the Clamshell Protocol will be reported by describing the frequency of adverse events (AEs) related to bone and superficial tissue healing in the postoperative period. It was hypothesized that the less-restrictive protocol will not lead to increased AEs. This will serve as a preliminary step toward establishing the safety of this new protocol following LTx and, ultimately, its efficacy.

MATERIALS AND METHODS

Study Design

This pilot study was a retrospective, observational, quasi-experimental study with cohorts based on time before and after implementation of the new protocol.

Setting

A single, academic, medical center with a transplant program in the intermountain west region participated in this study. Because this was a retrospective analysis of a clinical intervention, institutional review board (IRB) approval (IRB_ 00162175) was obtained, but patient consent was not required.

Participants

All adult patients who underwent bilateral LTx at University of Utah Health between March 1, 2021, and February 28, 2023, were eligible for inclusion. All bilateral LTx at the study center used clamshell incisions. Patients were excluded from the final analysis if they did not survive their initial hospitalization or if they did not participate in the outpatient PR programming. Cohorts were determined based on the postsurgical precautions in place at the time of their LTx. The sternal precaution (SP) cohort included individuals whose LTx occurred between March 1, 2021, and February 28, 2022, before implementation of the new Clamshell Protocol. The clamshell precaution (CP) cohort included individuals whose LTx occurred between March 1, 2022, and February 28, 2023, after the protocol implementation.

Movement Protocols

A novel LTx rehabilitation protocol, the Clamshell Protocol, was developed by PT staff members in conjunction with physicians from pulmonology and thoracic surgery. The Clamshell Protocol consisted of reduced postsurgical UE movement precautions and a rehabilitation program incorporating more UE resistance training. Full details of the new protocol, including weekly progressions, are described in our previous work.¹⁶ A summary of the SP and CP UE movement restrictions and rehabilitation exercises are summarized in Table 1. Although compliance with movement restrictions was not formally tracked, it was standard of care to assess and address compliance with both UE movement restrictions and exercise programming at each PT or PR session.

TABLE 1.

Comparison of Sternal Precaution (SP) and Clamshell Precaution (CP) Protocols

	SP Protocol	CP Protocol
Allowed UE movements	Non–weight-bearing and close-to-body movement only	Close-to-body functional movements even if weight-bearing, trunk rotation, stretching in pain-free range
Weight restrictions	5 lbs or less	5 lbs or less for first 4 weeks, progressing weight limit by 5 lbs each week
Time frame of precautions	8 weeks	Progressing after week 4, lifted after week 8
Initiation of basic mobility training	Within 24 hours or as soon after LTx as medically appropriate	Within 24 hours or as soon after LTx as medically appropriate
Initiation of pulmonary rehabilitation protocol	At time of hospital discharge	At time of ICU discharge
Exercise restrictions	No stretching, trunk rotations, or UE strengthening with more than 5 lbs for 8 weeks	No UE strengthening with more than 5 lbs for 4 weeks, progressing limit by 5 lbs per week

LTx, lung transplantation; lbs, pounds; 5 lbs, equivelant 2.3 kilograms; UE, upper extremity.

Variables

All variables were electronically abstracted from the electronic medical record. The primary outcome was AEs. AEs were identified using International Classification of Diseases Revision 10 (ICD-10) codes.¹⁷ After collecting all ICD-10 codes for each unique patient from day of LTx and onward, analysis focused on AEs specifically related to the surgical incision and sternum. Identified ICD-10 codes used at our institution included those found below in Table 2. These ICD-10 codes correlate with any medical code for initial presentation or follow-up presentation of infection, fracture, disruption, dehiscence, or nonunion of any surgical wound or surgically repaired bone.¹⁷ To focus on AE proximity to the intervention, analysis was further narrowed to AEs occurring within 6 months of the LTx date. All abstracted AEs underwent manual chart verification. Adverse events were subsequently categorized as either infection related or bone healing related. Additional details from physician notes related to the AE, days on mechanical ventilation, hospital length of stay, and time (in days) from surgery to when the AE code was first documented were abstracted. At this time, age, weight, body mass index (BMI), and the Charleston Comorbidity Index (CCI) at the time of LTx were also abstracted, as these were previously shown to correlate with AE risk in surgical patients.^18,19

TABLE 2.

Description of ICD-10 Codes Used for Identification of an Adverse Event (AE)

ICD-10 Codea	Code Description	AE Category
T81.31	Disruption of external operation (surgical) wound	Bone
T81.32	Disruption of internal operation (surgical) wound	Bone
T81.49	Infection following a procedure, other surgical site	Infection
S22.2	Fracture of the sternum	Bone

^aThe selected codes were determined to encompass all codes related to AEs of interest for this study. Code descriptions taken from ICD-10 Codebook. Codes shown with only diagnostic categories and not specific etiology or severity codes. Categories of bone, referring to body healing complication, or infection, referring to infectious process leading to wound complications given by researchers.

ICD-10, International Classification of Diseases Revision 10.

Statistical Analysis

Demographics and clinical characteristics were described in mean and standard deviation (SD) or counts and percentages. They were then compared between the 2 cohorts using Fisher exact test and Welch t test with unequal variances for categorical and continuous variables, respectively. To assess the safety of the new protocol, the number and type of AEs overall and in each cohort are described, along with 95% binominal confidence intervals (CIs) calculated using the Wilson score method. Clinical characteristics of the individuals who did and did not experience an AE at any record in each cohort were also summarized. We chose not to perform statistical comparisons on the characteristics between those with and without AEs because AEs were rare, and this was pilot work. Statistical analyses were conducted in R v.3.4, significance was assessed at the 0.05 level, and all tests were two tailed.²⁰

RESULTS

Forty-six bilateral LTx surgeries were performed at the study center during the two-year study period. After exclusion of patients who died before initial discharge (n = 1) and those who did not receive PR interventions (n = 5), 40 patients remained for analysis (16 in the SP cohort; 24 in the CP cohort; Fig. 1). There was no significant difference in baseline demographics or clinical characteristics between the 2 groups (Table 3).

Fig. 1.

CONSORT diagram. This diagram described the inclusion and exclusion of participants and the final number of participants in each study cohort.

TABLE 3.

Summary of Baseline Demographics and Clinical Characteristics of Study Cohorts

Variable	Total (N = 40)	SP (N = 16)	CP (N = 24)	P
Age, mean (SD)	59.7 (9.3)	59.7 (9.6)	59.7 (9.3)	1.00a
Sex, male n (%)	31 (77.5%)	13 (81.2%)	18 (75.0%)	0.72b
Female n (%)	9 (22.5%)	3 (18.8%)	6 (25.0%)	—
Race, n (%), White	35 (87.5%)	12 (75.0%)	23 (95.8%)	0.14b
Other	5 (12.5%)	4 (25.0%)	1 (4.2%)	—
Ethnicity n (%), Hispanic/Latino	2 (5.0%)	2 (12.5%)	0 (0.0%)	0.15b
Not Hispanic/Latino	38 (95.0%)	14 (87.5%)	24 (100.0%)	—
BMI, mean (SD)	25.6 (4.5)	25.0 (5.2)	26.0 (4.1)	0.56a
Weight (kg), mean (SD)	77.9 (16.5)	75.6 (18.3)	79.4 (15.5)	0.50a
CCI score, mean (SD)	2.7 (1.7)	2.6 (1.5)	2.8 (1.9)	0.67a
Clinical characteristics, mean (SD)
Hospital length of stay after LTx (days)	18.7 (9.1)	18.1 (11.7)	19.1 (7.2)	0.77a
Number of pulmonary rehabilitation sessions	28.3 (9.5)	29.7 (8.5)	27.3 (10.1)	0.43a
Pulmonary rehabilitation weeks	17.4 (15.1)	17.8 (9.3)	17.2 (18.1)	0.90a

All demographics reported reflect those at time of LTx surgery. All clinical characteristics include health care use after time of transplant only.

^aWelch t test.

^bFisher exact test.

BMI, body mass index; CP, clamshell precaution; CCI, Charlson Comorbidity Index; kg, kilograms; LTx, lung transplantation; N, total number; n, number; SD, standard deviation; SP, sternal precaution.

Initially, 12 AEs occurring within 6 months of the LTx date codes were identified for 6 people (Table 4). Upon manual review, 2 of the individuals were determined to not have had a true AE. The first, number 5 below, had the codes placed in error on the day of transplant surgery. The second, number 6 below, had an unrelated cardiac surgery requiring a redo sternotomy after the LTx. Removing these 2 encounters, there were 4 patients (10%) with AEs (95% CI: 6.0%–26.9%), including 3 (18.8%) for the SP cohort (95% CI: 6.6%–43.0%) and 1 (4.2%) for the CP cohort (95% CI: 0.7%–20.2%). No statistical comparison was made due to the rarity of AEs; however, there were fewer AEs in the CP group, providing preliminary evidence as to the safety of the new protocol. Characteristics of individuals who experienced an AE appeared to be older, have increased BMI and weight, and have higher CCI scores compared with those who did not experience AEs as seen in Table 5.

TABLE 4.

Descriptive Details for Adverse Event Codes: Summary of adverse event (AE) Codes Found and Details Pertaining to That Code Including the Specific ICD-10 Code, AE Class (Bone or Infection), and for Each Patient With an AE, the Time on mechanical ventilation (MV), Hospital length of stay (LOS) and the Time From lung transplantation (LTx) to the Initial Coding of an AE

Unique Patient	Cohort	ICD-10 CODE	AE Class	Details	Time on MV (days)	Hospital LOS (days)	LTx to AE (days)
1	SP	T81.31XA	Bone	Infected chest wall requiring reopening of surgical site for washout	5	70	37
		T81.49XA	Infection				37
2	SP	T81.31XA	Bone	Ground level fall from orthostasis resulting in dehiscence of sternum, requiring plating. After initial sternal plating, patient incurred another fall and required a second sternal plating surgery	2	10	96
		T81.32XA	Bone				96
		T81.32XA	Bone				59
		S22.20XD	Bone				67
		T81.49XA	Infection				96
3	SP	T81.31XA	Bone	Superficial clamshell wound incision, sternal displacement and mediastinal and pleural infection. Required sternal plating	1	29	34
		T81.32XA	Bone				161
4	CP	T81.32XA	Infection	Sternal infection requiring washout and debridement, sternal rewiring and plating	1	13	37
			Bone
5a	CP	T81.49XA	N/A	AE codes placed day of surgery in setting of initial sternal incision and COVID-19 infection. Not considered an AE.	N/A	N/A	0
		T81.31XA					0
6a	CP	T81.31XA	N/A	Patient had redo sternotomy in setting of cardiac surgery unrelated to previous transplant surgery. Not considered an AE.	N/A	N/A	72

^aThe events related to placement of the codes for individuals 5 and 6 were determined to not be a true AE related to the LTx.

CP, clamshell precaution; ICD-10, International Classification of Diseases Revision 10; LTx, lung transplantation; SP, sternal precaution.

TABLE 5.

Summary of Demographics of Interest Stratified by Cohort and Incurrence of Adverse Events (AEs)

Variable; Mean (SD)	SP		CP
	AE (N = 3)	No AE (N = 13)	AE (N = 1)	No AE (N = 23)
Age, yr	63.4 (6.7)	58.9 (10.1)	60.4 (NA)	59.7 (9.5)
BMI	27.7 (6.2)	24.4 (5.0)	25.2 (NA)	26.0 (4.2)
Weight (kg)	90.4 (16.5)	72.1 (17.4)	79.7 (NA)	79.4 (15.9)
CCI score	3.7 (2.1)	2.3 (1.3)	3.0 (NA)	2.8 (2.0)
PR sessions	31.7 (5.5)	29.2 (9.2)	36.0 (NA)	27.0 (10.2)

All demographics reported reflect those at time of LTx surgery. PR session describes the number of sessions after LTx only. P values are not provided due to the small sample sizes.

BMI, body mass index; CP, clamshell precaution; CCI, Charlson Comorbidity Index; kg, kilograms; N, number; N, total number; PR, pulmonary rehabilitation; SP, sternal precaution; SD, standard deviation.

All ICD-10 codes shown with specific etiology or details as coded in the electronic medical record and AE class as identified by initial category in Table 1.

DISCUSSION

This study preliminarily examined the safety of reduced postsurgical UE movement restrictions after bilateral LTx using a clamshell incision. Although AEs were rare, data demonstrated no increase in AEs related to healing in the CP group compared with the SP group. This preliminary analysis aligns with the evidence supporting less-restrictive precautions for median sternotomy incisions and adds to the growing body of literature examining whether postsurgical protocols may be too restrictive.^10-12,14

Postsurgical precautions are designed to minimize the incidence of AEs, but it is crucial to balance this risk with the desire to improve function, HRQOL, and pain. Researchers and clinicians must evaluate the efficacy of precautions for each unique type of surgical incision. Presently, very few studies have studied restrictions following surgeries with clamshell incisions or any specific to LTx. In addition, research and clinical practice should consider if the precautions may be negatively affecting patient's pain, fall risk, and HRQOL, as these are of particular interest in this high-risk surgical population. In a novel study by Fuller et al,¹⁴ adding UE exercises and stretching during PR programming following LTx improved reported pain and HRQOL without any noted change to other surgical outcomes. An additional study by Tarrant et al²¹ noted that starting more intensive rehabilitation in the inpatient setting was safe and feasible—a finding that was corroborated by the early start structure of our study.

A close look at the incidences of AEs revealed that many are related to infections of various sources or traumatic falls, factors likely unrelated to UE movement precautions. The concern that strength training, flexibility, or other exercise involving the upper extremities in the postsurgical period may cause AEs should continue to be investigated given the findings in this pilot data. Patient demographics, such as age, weight, or CCI score, may be more closely correlated with wound dehiscence or other incision complications. Given this, a stratified approach to allow for more conservative precautions for those who possess high-risk characteristics may also warrant consideration as has been suggested in sternal precaution literature.²²

Individuals undergoing LTx are more likely to be older, frail, chronically deconditioned, and have a longstanding and ongoing need for corticosteroids and immunosuppressants compared with those undergoing many other types of surgery.²³ They also, importantly, are at high risk for movement restrictions affecting their function and HRQO due to size and location of the surgical wound, the need to use UE support for mobility devices, and higher than average fall risk.⁷ These risks are what led to the creation of the Clamshell Protocol.¹⁶

Limitations

The Clamshell Protocol was implemented as a new standard of care for all LTx surgeries occurring at the medical center. This prohibited a randomized trial design. The observational and retrospective nature of this study does limit the overall quality and ability to control for confounding variables. As this was a pilot study, small sample size and low statistical power inhibit generalization at this time.

Future Directions

Further research in the safety, efficacy, and implementation of this novel Clamshell Protocol will be important in determining if such a protocol may be an improved standard of care following LTx. A larger, prospective, randomized trial or multisite trial using the same protocol would provide more robust data to confirm this.

This pilot study was developed to include information about functional and HRQOL outcomes for LTx recipients as such outcomes were the driving force behind the creation of the Clamshell Protocols. Because of resource strain resulting in inconsistencies in clinicians obtaining and recording these objective measures, this information was not able to be included in this analysis. The objective measures designed in the Clamshell Protocol remain the ideal outcomes for tracking the impact of this new protocol, as it relates to physical function, dyspnea, pain, and HRQOL, and should be collected and analyzed in future research.¹⁶

CONCLUSIONS

A novel rehabilitation protocol to reduce UE movement restrictions following bilateral LTx using a clamshell incision was successfully implemented at a single academic medical center. Pilot data from this study provided preliminary evidence of reduced AEs related to the healing of surgical wounds, including clamshell incisions. Increased evidence to support the Clamshell Protocol through future research may support its implementation in LTx programs nationally to reduce restrictions and improve outcomes for individuals in this important surgical population.

Clinical Pearls

• A new Clamshell Protocol for upper extremity movement precautions following lung transplantation was developed by the study team. See publication Slay et al (2025) for details.

• The safety of implementing the new protocol was assessed using a pilot retrospective cohort study to look for adverse events related to healing of the surgical incision following lung transplantation.

• There was preliminary evidence of lower adverse events due to implementation of the new protocol, providing preliminary support of its safe use in this population.

• Further research with larger populations is needed to statistically analyze safety as well as the protocols impact on patient function and health-related quality of life.