A Claims Based Natural History of the Early Postoperative Period after a Unilateral
Shoulder Rotator Cuff Repair
Stephan Pill MD, MSPT, FAAOS1, Samantha J Beckley PhD2, Maha Karim BS2, Shaun K
Stinton PhD2, Thomas P Branch MDb
1 Prisma Health Department of Orthopedic Surgery, Steadman Hawkins Clinic of the Carolinas,
Greenville, SC, USA.
2 ArthroResearch LLC, 441 Armour Place NE, Atlanta, GA, 30324, United States
3 Ermi LLC, 2872 Woodcock Blvd. Suite 100, Atlanta, GA, 30341, United States
Address for correspondence:
Samantha Beckley, 441 Armour Place NE, Atlanta, GA, 30324, USA
Phone: +1 404-579-1546
Email: s.beckley@arthroresearch.com.
|
Abstract:
Background: The aim of this paper was to determine a claims-based natural history of the
postoperative period after a unilateral shoulder rotator cuff repair including the impact of
comorbidities and post-surgical complications on recovery.
Methods: Healthcare claims data were analyzed to evaluate costs and recovery timeline
following unilateral shoulder rotator cuff repair. The analysis included costs associated with: i)
index surgery, ii) complications requiring repeat surgery, iii) conversion to shoulder arthroplasty,
iv) non-operative hospitalization, v) motion restoring procedures and vi) shoulder-related
outpatient surgeries. The effect of comorbidities was determined using data from patients with
diabetes, obesity, peripheral vascular disease, and cardiovascular disease. The effect of
post-index surgical complications was determined using data from patients who were
rehospitalized (with or without additional surgery). Perioperative complications including joint
contracture/adhesive capsulitis, infection, and pulmonary embolus were also examined.
Results: Index surgery median cost was $11,454 (IQR:$8,169-$17,204). Median length of
postoperative recovery (from Index Surgery to last physical therapy claim) was 153 days
(IQR:79-683 days). A total of 46% patients took longer than six months to complete their
postoperative course of physical therapy. The development of a shoulder contracture or adhesive
capsulitis in the postoperative period added a median of 162 days to recovery and nearly double
the normal postoperative cost. The recovery period and cost for patients that required surgery for
a complication were 3.5 times longer and 5 times the cost in comparison to patients not requiring
any complication surgeries. Co-morbidities added 30 to 90 days to recovery.
|
|
Conclusion: Comorbidities and complications following unilateral rotator cuff repair contribute
to considerable increases in costs and prolonged recovery periods. Knowing the average timeline
of recovery in patients after rotator cuff repair surgery and factors that increase the length in
postoperative therapy services may help become a surrogate for success of treatment, predict
outcomes, and manage costs and value.
Keywords (6-8 required): Shoulder; Contracture; Motion loss; Claims analysis; Rotator cuff
repair; Postoperative recovery
|
|
Introduction:
Understanding and defining the current common postoperative course after a surgical
intervention is the first critical step in the process towards improving it. Most clinical studies on
surgical treatments are reported as a case series with a minimum of 2 year follow-up, resulting in
a limited amount of information available about the immediate postoperative period. However, a
significant amount of healthcare dollars are spent on recovery in the first year after a surgery.
Understanding recovery patterns in the postoperative period requires a substantially larger data
set than typically available to a standard case series in order to include all possible pathways as
outcomes.
Shoulder rotator cuff repair (RCR) is among the most commonly performed treatments for
shoulder pain in the US and has been increasing at a steady rate over the last 10 years 1
. Full
recovery from RCR surgery can be hindered by comorbidities or complications related to the
surgery or during recovery 2-13
.
In patients struggling with their recovery, treatment options include continued physical therapy
(PT) alone, injections, Motion Restoring Surgeries (MRS) such as manipulation under
anesthesia (MUA), arthroscopic lysis of adhesions and/or arthroscopic/open debridement21
.
Additional surgical interventions related to complications may include revision to arthroplasty or
arthroscopic interventions, such as revision repair, or repeat or added procedures involving other
structures, such as the biceps or acromioclavicular joint. All of these additional interventions can
lead to increased risks, costs, and delayed recovery 14
. A clearer understanding of the typical
resource utilization, variation, and incidence of complications and repeat surgeries during the
immediate post-operative course after rotator cuff surgery will help identify outliers and may
|
|
identify those who are falling off an acceptable post-operative course. Patients deemed as
laggards may constitute a large portion of healthcare spending related to rotator cuff surgery, and
attempts to direct these patients to other treatment alternatives may be more appropriate.
Moreover, early healthcare resource utilization may be a surrogate for eventual outcome.
Addressing the postoperative challenges may help develop protocols that reduce recovery
obstacles and enhance patient outcomes.
A useful tool for identifying RCR recovery patterns is the IBM MarketScan Commercial Claims
and Encounters Database. This resource provides health information from reimbursed healthcare
claims, submitted through employers and health plans 15
. It includes comprehensive details on
clinical utilization, healthcare costs, insurance coverage, and care delivery for approximately 30
million individuals per year in the United States.
This study aimed to analyze claims data to construct a timeline for recovery following unilateral
RCR, covering the period from the initial surgery to the final PT session. Furthermore, the study
aimed to quantify the effects of comorbidities and post-surgical complications on cost and
timeline recovery.
Methods:
Study Population
This study utilized healthcare claims data from the IBM MarketScan database (2015–2018),
which includes information on 11 million unique patients and 1 billion claims. This retrospective
study was determined to be exempt from IRB review. Eligibility criteria required patients to
have maintained continuous insurance coverage for at least two consecutive years. Patients
|
|
having RCR surgery were identified by CPT codes 23410, 23412, 23420, and 29827. Those with
bilateral RCR surgeries, either simultaneous or staged, were excluded, as bilateral surgery would
confound the postoperative recovery period and skew the recovery both longer and with greater
expense. The postoperative period was defined as the timeline from the index RCR surgery to the
last Healthcare Common Procedure Coding System (HCPCS) CPT charge for PT. Only patients
with at least one PT CPT charge (codes >=97000 & <98000) post-surgery were included.
Cost Analysis
Costs associated with the index surgery were defined as all healthcare expenditures incurred
during the inpatient hospitalization or within eight days following an outpatient procedure (the
eight day period was necessary to capture all outpatient surgery costs). Subsequent procedures
were categorized into the following groups: (i) complication-related surgeries, (ii) revision to
arthroplasty or arthroplasty revisions, (iii) non-operative hospitalizations, and (iv)
motion-restoring surgeries. Total costs for each category included the initial index surgery cost
plus all subsequent shoulder related events associated with that category, eg. if a patient has a
MRS followed by a complication followed by an arthroplasty, all costs are included. Standard
CPT codes were used to analyze postoperative costs, including PT (>=97000 and <98000),
physician visits (>=99200 and <99300), injections (20610 and 20611), and radiology
(Addendum #1). Costs unrelated to the RCR were excluded from the analysis if shoulder-related
ICD-9/10 codes were absent on the HCPCS form. All reported costs reflect insurance-paid
amounts.
Comorbidities
|
|
Comorbidities were analyzed to determine if recovery timelines were affected. Patients with
diabetes (codes 25.0 & E11.9), obesity (codes 278., E66., & Z68.4), peripheral vascular disease
(PVD, codes 440.:444., 785., & I73.9), joint infection (codes 711., 996., M00., M01, & M02),
cardiovascular disease (CVD, codes 390.:459., I11, I20, I21, & I25), and shoulder contracture
(ICD 9/10 codes 718.41, M24.51, 726.0, M75.0). Median costs and interquartile ranges (IQR)
were calculated for each group. Infection (Addendum #6) and pulmonary embolus (Addendum
#7) incidence rates associated with each comorbidity were also included.
Index Surgery
All expenses related to the initial surgery were identified using the inpatient surgery CaseID. For
surgeries performed in an outpatient setting (without a CaseID), all costs incurred within an
eight-day period were included in the total index costs, as claims were submitted over time. If the
RCR procedure began in an outpatient setting but the patient was transferred to a hospital,
generating a new CaseID within eight days, these costs were included in the outpatient index
surgery expenses. Patients who underwent inpatient surgery and were later transferred to an
inpatient rehabilitation facility were counted as having two separate hospitalizations. The
inpatient rehabilitation stay was classified as a nonoperative hospitalization.
Additional Postoperative Surgeries
The influence of post-surgical complications on recovery was examined, with particular attention
to rehospitalizations and repeat surgeries, such as revision to arthroplasty (Addendum #2), MRS
(Addendum #3), and other surgeries related to complications (Addendum #4). Hospitalizations
were identified through a new CaseID associated with or occurring after the initial surgery.
|
|
Perioperative complications studied included joint contracture (Addendum #5), infections
(Addendum #6), and pulmonary embolism (Addendum #7).
Presentation of Recovery Pathway
An ideal recovery pathway assumes a smooth postoperative course, where the patient undergoes
the initial surgery, is discharged timely, participates in PT, and achieves full recovery within six
months. It is widely recognized that an injury or surgery transitions from acute to chronic after
six months 16
, with many states and insurance providers limiting temporary or short-term
disability coverage to this timeframe. To demonstrate the challenges of complex recoveries, a
multi-pathway chart (Figure 1) illustrates the combined effects of multiple postoperative events,
including Motion Restoring Surgeries (MRS), complication surgeries and revision to arthroplasty
surgeries.. Rather than representing a specific timeline, the chart highlights the cumulative
impact of successive complications during recovery.
Data Analysis
Data analysis was performed using the R statistical programming language (version 4.3.3).
Given the large sample size, results are presented as medians with IQR.
Results:
Study Group Demographics
The study group was composed of 14,947 patients, of which 57.7% were male. The median age
was 55 years (11-63 years). Data for patient height and weight were not available.
Index Surgery Costs
|
|
Of the 14,947 surgeries, 14,919 were outpatient, 26 were inpatient, and two started as outpatient
and ended as inpatient. The index surgery median cost was $11,454 (IQR:$8,169-$17,204). The
median cost of inpatient surgery was $17,409 (IQR:$8,964-$31,294) while the median outpatient
surgery cost was $11,457 (IQR:$8,172-$17,200). The two patients that started outpatient and
ended inpatient had a median cost of $18,548 (IQR:$17,019-$20,078). Concomitant surgeries at
the time of the index surgery included biceps surgery in 43% of cases, labral surgery in 7%,
distal clavicle resection in 36%, and acromioplasty in 81%.
Recovery Period
The median length of postoperative recovery (from index surgery to final PT claim) for a
unilateral RCR was 153 days (IQR:79-683 days). Only 54% of patients completed their
postoperative period within the initial six months (See Figure 2). Patients who completed their
postoperative course in less than six months, spent a median of 83 days in structured outpatient
PT; whereas, patients who completed their postoperative course in over six months spent a
median of 719 days in structured outpatient PT. The additional postoperative costs incurred after
six months for patients undergoing PT beyond that point had a median of $2,065
(IQR:$1177-$4058).
Postoperative Complications
We examined the postoperative timeline and cost effects of four major events that may occur
after a unilateral RCR: 1) MRS, 2) Complication surgery related to the RCR, 3) Revision to
arthroplasty and 4) Nonoperative hospitalizations related to infection or pulmonary embolus.
After MRS, such as a MUA, arthroscopy for lysis of adhesions, debridement, or synovectomy,
both the median number of days in structured PT and the costs of postoperative care nearly
|
|
tripled (Table 1). The median recovery period and cost for patients that required a surgery for a
postoperative complication were 3.5 times longer and added five times the cost in comparison to
patients that did not require any complication-related surgeries (Table 1). Revision surgery had
the same impact (Table 1). Nonoperative hospitalizations for pulmonary embolus and/or
infection increased the postoperative period nearly four times (607 days vs. 153 days) at a cost of
a median $31,113 per event (Table 1). The cumulative effect on medical care costs and recovery
time as a result of multiple events in the postoperative period can be seen in Figure 3 and Table
2. In particular, the decision to perform a MRS substantially impacted the risk of requiring a
shoulder salvage procedure, with patients undergoing MRS being seven times more likely to
have a revision to shoulder arthroplasty compared to those without a MRS (0.7% vs. 0.1%).
There were a significant number of repeat procedures in the postoperative period from the same
set seen at index surgery, including distal clavicle resection, biceps, acromioplasty, labrum
related procedures (Table 3).
A single RCR with no complications, no revisions and no MUA had a median postoperative cost
of $2,708. We found that 8.7% of single RCRs required a secondary surgery within the defined
postoperative period. The median cost surged dramatically to $41,880 with a revision to a
shoulder arthroplasty. Out of the 14,947 patients that were observed with a unilateral RCR,
14,109 patients did not undergo a MRS. These patients had a median of 145 PT days and the
median was $2,806. Of the 838 patients who did have MRS, the median cost was $17,352
(IQR:$10,043-$27,345) with a median of 451 PT days.
Shoulder Joint Stiffness
|
|
The presence of joint fibrosis, which was defined as a shoulder contracture or adhesive
capsulitis, at the time of index surgery did not affect recovery costs or the time required for
recovery. However, the onset of joint fibrosis following the initial surgery significantly
influenced both. The median cost and recovery time for the subset of patients that developed a
joint fibrosis postoperatively was twice as expensive and two times longer than the subset of
patients with who did not develop fibrosis (Table 4).
Impact of Comorbidities on Recovery
Table 5 illustrates the impact of diabetes, obesity, PVD, and CVD. The influence of these
conditions on recovery outcomes was minimal. However, patients with PVD and CVD
experienced a higher frequency of pulmonary embolus events compared to those without these
comorbidities.
Discussion:
The main finding of this study was the establishment of a benchmark for the postoperative
recovery period following unilateral RCR whereby patients were found to have large variations
in cost and healthcare resource utilization. The median recovery time was 153 days, with nearly
half of the patients requiring more than six months to complete their recovery. This finding
matches a previous study that reported a significant number of patients who experienced delayed
recovery as shown by 28% of patients failing to reach functional recovery within 6 months 17
.
These large variations in recovery timelines highlight the complexity of the RCR recovery
process.
|
|
Another critical finding was the substantial impact of major postoperative events such as MRS,
complication-related surgeries, revision surgeries, and non-operative hospitalizations on both
recovery time and healthcare costs. The need for revision surgery after a unilateral RCR included
repeating the rotator cuff repair in some cases or doing other “adjunctive” surgeries not
performed at the index surgery, such as distal clavicle, biceps, labrum, and/or acromion
procedures. Delaying these procedures until after the index surgery significantly increased
overall costs.
Truong et al. studied reoperation rates in a large database of 24,392 patients who had 2-year
follow-up data and found a 10.4% overall reoperation rate with 1.3% of those operations being
conversion to shoulder arthroplasty. They also compared patients who underwent open RCR to
patients with arthroscopic repair and found greater rates of 30-day emergency department
(7.0%), 30 day hospital readmission (2.0%), and subsequent surgery (11.3%) in open RCR
compared with arthroscopic RCR (6.3%, 1.0%, 9.5% respectively) 18
. Patients experiencing
complications incurred the highest financial burden, and those requiring revision surgery faced
the longest recovery durations. These results are similar to our findings and emphasize the
financial and temporal toll of postoperative complications and underscore the need for strategies
to reduce these risks.
Another significant observation was the effect of postoperative shoulder fibrosis, defined as
contracture (M24.51, 719.51) or adhesive capsulitis (M75.0, 726.0), on recovery outcomes. The
development of fibrosis after surgery led to a marked increase in both costs and recovery time.
However, the presence of fibrosis prior to surgery did not influence claims-based outcomes.
Patients who developed fibrosis postoperatively incurred significantly higher median costs and
longer recovery times than those without fibrosis. Previous studies have reported a significant
|
|
incidence of postoperative stiffness or shoulder contracture with an incidence ranging from 4.9%
to 32% 3,4,6,7,11-13
. The delay in recovery due to postoperative shoulder contracture can still exist at
one year followup 8,11
. In this study, we found that 8.7% of single RCRs required a secondary
surgery which is slightly higher than in a previous study where 4.9% of arthroscopic RCR cases
required an additional arthroscopic capsular release 7
.
Comorbidities also had a notable impact on recovery time and costs. Patients with PVD,
diabetes, obesity, or CVD experienced an average delay of 54 days in their recovery. These
conditions, along with related adverse events, prolonged the recovery process and often required
additional medical interventions, extended rehabilitation, longer hospital stays, and specialized
care, all of which contributed to increased healthcare costs and recovery times 13
. Diabetes and
obesity have been previously shown to affect the speed of recovery after RCR and to increase the
risk of revision matching the results found in this study 5,9,13,19,20
.
This study has limitations inherent to HCPCS claims-based analyses. Standard coding
inconsistencies, particularly during the transition from ICD-9 to ICD-10 codes, may have
introduced variability. Additionally, the definition of the postoperative period assumes recovery
concludes with the final PT visit, which may not capture the full scope of medical care. Another
limitation is that the claims database did not include the cost of prescribed medications, which
would likely add to the total healthcare expenses. ICD coding for joint motion loss has not been
defined well enough for clinicians to apply easily, which may cause confusion with coding.
Joint ankylosis (M24.6x) is a fixed loss of joint motion. Joint contracture (M24.5x) is non-fixed
or recoverable joint motion loss. Joint stiffness (M25.6x) is a difficult to move joint with no
motion loss. Adhesive capsulitis is motion loss due to inflammation (M75.0).
|
|
Conclusion:
The benchmark established in this study provides a standardized framework for tracking and
managing patient progress during the early stages of recovery. This timeline is essential for
evaluating the impact of new strategies designed to enhance recovery outcomes. Without this
foundation, the effectiveness of new approaches can only be measured through randomized
controlled trials. Additionally, the significant recovery delays of nearly half% of patients taking
over six months to recover highlight the need for improved postoperative care protocols for
unilateral RCR patients.
|