Brayton, K.M., Hirsch, A.T., O'Brien, P.J., Cheville, A., Karaca-Mandic, P., & Rockson, S.G. (2014). Lymphedema prevalence and treatment benefits in cancer: Impact of a therapeutic intervention on health outcomes and costs. PLOS One, 9, e114597. 

To examine the insurance data of cancer survivors to determine lymphedema prevalence and to assess the affect on specific clinical health outcomes and healthcare costs after the receipt of a pneumatic compression device (PCD) among the identified study group of cancer survivors with lymphedema

The researchers performed a retrospective analysis of health claims data from 2007–2013. D-identified administrative health claims data from deidentified Normative Health Information (dNHI) database (34 million insured) were used through OptumInsight. Researchers did not have access to the data but rather instructed Optum employees to cull the data.  

• ICD-9 codes were used to select patients with cancer.
• ICD-9 codes for lymphedema (457.0, 457.1, and 757.0) were sought among the patients with cancer.  
• The Healthcare Common Procedure Coding System (HCPCS) codes for a PCD, E0651, or E0652 were searched among patients with cancer and lymphedema.
• Every participant was required to have had 12 months of insurance coverage prior to the receipt of a PCD, which then carried through another 12 months.  
• Any patients who received a replacement pump were excluded.

• N = 1,065   
• AGE = 9.2% aged 19–44 years, 53.2% aged 45–64 years, and 37.6% aged 65 years or older
• MALES: 20%, FEMALES: 80%
• CURRENT TREATMENT: Not applicable
• KEY DISEASE CHARACTERISTICS: Patients with cancer with insurance for 12 months prior to and after the receipt of a PCD (index date). Baseline period was 12 months prior to receipt of the pump. Follow-up was 12 months after receipt of the pump.  
• OTHER KEY SAMPLE CHARACTERISTICS: Nineteen percent were obese.

• SITE: Multi-site   
• SETTING TYPE: Not specified    
• LOCATION: USA

A retrospective study of insurance data from 2007–2013. The researchers compared the rate of identified clinical healthcare outcomes and their costs in each setting for the year prior to the receipt of a PCD to the rates and costs after receipt of the PCD.

After the sample was identified, the affect of the receipt of the PCD was determined by searching for specific claims codes and their costs during the 12 months before and after receipt of the PCD. Specific health outcomes, such as hospitalization, outpatient visits, physical therapy, episodes of cellulitis, and courses of lymphedema physical therapy, were determined by the American Medical Association place of service codes, and the clinical costs were designated as lymphedema related if the primary or secondary diagnoses were ICD-9: 457.0, 457.1, 757.0. Total cost was the sum of the payments for lymphedema claims. Continuous variables were tested pre PCD minus post PCD with a paired t test.

• Prevalence: In 2007, a prevalence of lymphedema cancer survivors existed at 0.95% in the database. The prevalence rose to 1.243% in 2013.
• Health outcomes: While receipt of a PCD appeared to have no affect on the rate of hospitalizations for patients with lymphedema, it did appear to reduce the percentage of patients attending outpatient hospital visits (p < 0.0001) and the percentage of patients using physical therapy (p < 0.0001). In addition, the number of patients diagnosed with cellulitis decreased significantly (p < 0.0001).  
• Healthcare Costs: The total healthcare cost for patients with cancer-related lymphedema during the baseline year was $62,190 with inpatient costs accounting for $15,458, outpatient costs accounting for $21,222, and office visits accounting for $15,278. However, lymphedema-attributed costs were only 4% of the total, or $2,243. A significant decrease in total cost by 18% occurred in the 12-month period after receipt of the PCD (from $62,190 to $50,857) (p < 0.0001).

• A study of the sample database showed an increasing prevalence of cancer-related lymphedema.
• The healthcare outcomes data showed a reduction in the rate of hospitalizations, outpatient visits, physical therapy, and cellulitis as well as a significant healthcare cost reduction the year following the receipt of a PCD.

• Risk of bias (no control group)
• Findings not generalizable
• The stages of cancer are not known or accounted for, and some of the costs may reflect the burden of cancer care. 
• The burden of disease is greater at the onset of lymphedema than when the limb swelling stabilizes. Likewise, the reduction of physical therapy reflects perhaps that the patients had physical therapy and then received a PCD for maintenance. Having a pump does not necessarily mean patients use the pump.
• It is unknown if the patients received a compression garment. 
• Other codes could have been looked at, including 97140, manual lymph drainage. 
• Too many confounding variables exist, making the results nongeneralizable. 
• As the authors pointed out, inaccuracy in coding cannot be ruled out. The low obesity/overweight rate of 19% was below the national average, which is currently 68.8%.

Lymphedema prevalence among cancer survivors continues to be defined. Whether a PCD improved healthcare outcomes or reduced healthcare costs is unclear. Nevertheless, healthcare outcomes and costs were reduced. A PCD may benefit outcomes and reduce costs. Even with the limitations of this study, lymphedema needs to be diagnosed, treated early, and managed to reduce the disease burden.