Abstract
Objective:
To assess the demographic and clinical characteristics associated with initiation of duloxetine therapy compared with other pharmacologic therapies for treatment of diabetic peripheral neuropathic pain (DPNP).
Methods:
This study used administrative claims databases for commercially-insured individuals aged 18–64 years to examine factors associated with treatment among DPNP patients who initiated duloxetine versus tricyclic antidepressants (TCAs), venlafaxine, gabapentin, pregabalin, or opioids between 7/1/2005 and 12/31/2007. Treatment initiation was defined as no pill coverage of the same medication over the previous 90 days. Multiple logistic regression models were estimated to assess factors associated with initiating duloxetine versus each of the other DPNP therapies.
Results:
The study included 11,060 DPNP patients with average age of 55 years old. Cardiovascular disease (63–70%), cerebrovascular/peripheral vascular disease (26–33%), low back pain (24–39%), and osteoarthritis (17–26%) were the most common diabetes- and pain-related comorbidities. Controlling for demographic and clinical characteristics, patients who received duloxetine or pregabalin in the prior 12-month period were more likely to initiate duloxetine. Patients from other DPNP treatment cohorts, except for those in the pregabalin cohort, were more likely to re-initiate the same prior therapy than begin treatment with duloxetine (all p < 0.05). A history of anxiety disorder was significantly associated with initiation of all DPNP treatments other than duloxetine (all p < 0.05), except for TCAs. Patients with low back pain were more likely to initiate duloxetine than TCAs or venlafaxine, but less likely to initiate duloxetine than pregabalin or opioids. Patients with infections related to diabetes were less likely to initiate duloxetine than venlafaxine, gabapentin, or opioids.
Limitations:
Because a retrospective administrative claims database was used, this study is subject to selection bias due to unobservable confounders, inability to measure prescriber preferences or characteristics or disease severity.
Conclusions:
Among commercially-insured DPNP patients, those with prior use of duloxetine or pregabalin were more likely to initiate duloxetine than other treatments. The presence of select comorbidities was also associated with specific medication initiation.
Introduction
Diabetic peripheral neuropathic pain (DPNP) is defined as spontaneous pain, or hypersensitivity to pain, caused by diabetic damage to the central nervous systemCitation1,Citation2. It is a relatively common complication of diabetes causing significant lifetime displeasure and morbidityCitation3. Up to half of all patients with diabetes have some degree of DPNP, with 10–20% experiencing moderate or severe painCitation3–8. Most DPNP patients experience pain on a daily basisCitation9. DPNP patients with chronic pain conditions are at increased risk of having comorbid mood, anxiety, and sleep disordersCitation10,Citation11. Not only do DPNP patients have impaired health-related quality of life, they also suffer from reduced functioning and increased disabilityCitation11–16.
Pharmacologic treatment options for DPNP management include tricyclic antidepressants (TCAs), anticonvulsants, serotonin-norepinephrine reuptake inhibitors (SNRIs), opioids, and topical and oral analog agentsCitation10. TCAs are widely used to treat chronic pain with low costCitation10; however, their use in clinical practice today is limited by an unfavorable tolerability profile and many contraindications which often lead to patient non-adherenceCitation10,Citation17. Duloxetine, an SNRI, has demonstrated safety and efficacy for managing DPNP and was approved by the US Food and Drug Administration (FDA) in late 2004Citation18,Citation19. Venlafaxine is another SNRI agent that appears to be effective for pain relief associated with DPNP, but is considered second-tier due to the lack of published dataCitation7. Anticonvulsants are also commonly used to alleviate DPNP symptomsCitation20, and pregabalin received its approval for DPNP indication from the US FDA in 2005. Opioids are another class of pharmacologic treatment available for DPNP management, with oxycodone and tramadol being the most commonly used opioidsCitation4. Opioids are generally used as adjunctive therapies, although there is no clinical trial evidence for this approachCitation17. In addition, the extensive list of side-effects and dependence issues are major concerns associated with opioid utilizationCitation21–23.
Selecting a medication can be challenging in the absence of head-to-head clinical trials for the numerous DPNP therapies. Choosing the appropriate treatment is highly individualized and largely based on trial and errorCitation24,Citation25. In addition to effectiveness, adverse events and patient comorbidities are central considerations for prescribing DPNP medications. Physicians should select a medication and then proceed through a stepwise progression of the various available therapies until a satisfactory level of pain relief is achieved without intolerable side-effectsCitation17,Citation25.
The objective of this study was to investigate the profile of DPNP patients who are more likely to be selected for duloxetine use versus other DPNP treatments using a large, US administrative claims database. No previous study has examined this since the FDA approval of duloxetine for DPNP, and no known research has explored whether there are systematic differences in the selection of patients for various DPNP pharmacologic treatments. If such differences exit, they could be helpful to physicians and patients in selecting the appropriate pharmacologic therapy. The knowledge gained from this study will be useful in generating hypotheses for future clinical research.
Methodology
Data source
This analysis utilized data from the Thomson’s Medstat MarketScan Commercial Claims files between July 1, 2004 and December 31, 2007. The database includes enrollment records, medical claims, and prescription claims of insured employees and their dependents from approximately 45 large health plans in the US. The enrollment data is a patient-level file that contains enrollment information such as age, gender, US geographical region, health plan type, employment status, as well as monthly enrollment status. The medical claims include inpatient and outpatient facility and services records. For service records, up to 15 diagnosis and procedure codes of International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) are reported. Each medical claim also includes date of service, place of service, provider type, and plan and patient paid amounts. The pharmacy claims record each prescription filled with information including the National Drug Code (NDC), dispense date, quantity and days supplied, and plan and patient paid amounts. All the information in the files can be linked via an encrypted and unique patient identification number. Because data used in this study were purchased from a third party which had removed identifying information prior to its release, institutional review board (IRB) and similar approvals were neither needed nor sought.
Sample selection
We identified patients who filled a prescription for any DPNP treatment, including duloxetine, TCAs, venlafaxine, gabapentin, pregabalin, and opioids (e.g., tramadol, oxycodone, morphine, hydrocodone, methadone, levorphanol)Citation7, between July 1, 2005 and December 31, 2007 (‘screening period’). The most recent initiation of therapy was selected, with an initiation defined as no pharmacy pill coverage for the same pharmacologic agent during the previous 90 days. For patients who had used the same medication over the prior 90 days, we searched for the next previous initiation. The corresponding dispense date of the most recent initiation was denoted as the index date, and the initiated medication was deemed the index agent. All eligible study patients were required to meet the following criteria: (1) continuous health plan enrollment over the 12 months before the index date, (2) age between 18 and 64 years as of the index date, (3) at least one prescription of pain medication (i.e., antidepressants, anticonvulsants, non-narcotic pain medications, opioids, non-steroidal anti-inflammatory drugs [NSAIDs], and topical and oral analogs) filled during the 12 months prior to the index date, and (4) at least one healthcare encounter for diabetes (ICD-9-CM: 250.xx) and DPNP (ICD-9-CM: 250.6x or 357.2) over the 12 months pre-index period. Six study cohorts were created based on the initiation of patients’ index agent. Individuals who initiated duloxetine, TCAs, or venlafaxine were excluded if they had a diagnosis of depression (ICD-9-CM: 296.2, 296.3, 300.4, 309.1, 311) in the 12-month pre-index period, while patients who initiated gabapentin or pregabalin were excluded if they had a diagnosis of epilepsy (ICD-9-CM: 345.xx, 780.39) or post-herpetic neuralgia (ICD-9-CM: 053.12, 053.19) in the prior 12-month period.
Study measures
Baseline demographic characteristics included age, sex, geographic regions (i.e., Northeast, North central, West, South), and health plan type (i.e., health maintenance organization [HMO], preferred provider organization [PPO], comprehensive, etc.). Patient copayments were reported based on the index record, and copayment of the index agent was standardized to a 30-day supply. Provider specialty (e.g., primary care providers, psychiatrists, neurologists, nephrologists, endocrinologists) was identified from the medical service claims occurring within 15 days prior to the index prescription. Clinical characteristics for the 12-month pre-index period included diabetes-related comorbiditiesCitation26–32 and pain-related conditionsCitation30–32 (see Appendix for details), as well as any use of insulin and/or oral anti-diabetic drugs (OADs) or pain-related prescription medications (i.e., antidepressants, anticonvulsants, non-narcotic pain medications, opioids, NSAIDs, and topical and oral analogs). All-cause, diabetes-, and DPNP-related inpatient utilization and healthcare costs were examined in the 12 months prior to the index date. Diabetes- and DPNP-related costs were extracted based on the medical service claims with such diagnoses coded in any of the diagnosis fields and prescription costs with associated medications recorded on the pharmacy claims. Diabetes-related medication costs were estimated based on prescriptions of insulin and oral anti-diabetic drugs (OADs). DPNP-related medication costs were estimated based on prescriptions of pharmacologic therapies recommended for DPNP treatment including TCAs, venlafaxine, duloxetine, pregabalin, gabapentin, and opioidsCitation10. All costs were adjusted to 2007 US dollars using the medical component of the consumer price indexCitation33.
Analysis
Descriptive statistics for demographics and clinical characteristics were summarized between the six cohorts. Percentages were reported for categorical variables and any differences between the duloxetine cohort and the other DPNP treatment cohorts were reported using chi-square tests, respectively. For continuous variables, mean and standard deviations were reported and Student t-tests were used to examine differences between the six treatment groups using the duloxetine cohort as the reference comparison group. Separate logistic regressions were estimated to assess the factors associated with initiating duloxetine versus each of the five other study mediations. For each model, the dependant variable was a dichotomous indicator coded as ‘1’ for patients in the duloxetine cohort and ‘0’ for patients in each comparison cohort. Two separate logistic regressions were estimated comparing duloxetine to each of the other alternative treatments, respectively: (1) explanatory variables included demographics, copayment associated with the index drug, prior use of pain medications, and prior use of anti-diabetic agents as a proxy for diabetes severity, and (2) explanatory variables included demographics, copayment associated with the index drug, diabetes-related comorbidities and pain-related conditions (as listed in the Appendix), and prior use of anti-diabetic agents. Adjusted odds ratios (ORs) and 95% confidence intervals (CI) were estimated. All analyses were conducted using SAS version 9.1 (SAS Institute, Inc., Cary, NC, USA).
Results
Demographic and clinical characteristics
This study included a final sample of 11,060 commercially insured DPNP patients, and and present the baseline characteristics for these patients by treatment cohorts. Approximately 33% of patients initiated therapy with gabapentin, 20% pregabalin, 18.5% opioids, 13.8% TCAs, 10.2% duloxetine, and 4.5% venlafaxine (). The average age of the study cohorts was around 55 years old, and patients in the TCAs and gabapentin cohorts were approximately 1 year older than those in the duloxetine cohort (both p < 0.05). Overall, 54% of all study patients were female, and the venlafaxine cohort had a higher proportion of females than the duloxetine cohort (66 vs. 56%, p < 0.05). Approximately half of all patients were geographically located in the southern region. Majority of patients visited a primary care provider prior to treatment initiation (69.3–81.8%).
Table 1. Baseline demographics.
Table 2. Baseline diabetes and pain-related comorbidities.
Common diabetes-related conditions () included cardiovascular diseases (prevalence ranged 63.4–70.2%), cerebrovascular/peripheralvascular diseases (prevalence ranged 26.1–32.7%). Among the six pain-related comorbidities we examined, the prevalence of low back pain ranged 24.0–38.5% and osteoarthritis 16.8–25.5%. A significantly higher proportion of pregabalin and opioid patients had cardiovascular disease than duloxetine patients (p < 0.05). Compared with patients in the duloxetine cohort, more patients in all other treatment cohorts had anxiety disorders. The proportion of patients with low pack pain was significantly higher in the pregabalin and opioids cohorts than the duloxetine cohort, while the prevalence was significantly lower in the TCAs, venlafaxine, and gabapentin cohorts.
Prior medication utilization
presents the medication use in the 12-month pre-index period. At least 70% of the patients in the TCAs, venlafaxine, gabapentin, and opioids cohorts had prior use of the index agent in the previous year excluding the prior 90-day pre-index period. In the duloxetine and pregabalin cohorts, 36.2% and 22.9% of patients had prior use of the index agent, respectively. Antidepressants and opioids were the most frequently prescribed medications among the duloxetine, TCAs, and venlafaxine cohorts, while opioids and anticonvulsants had the most use in the gabapentin, pregabalin, and opioids cohorts. Among the antidepressants, the use of the identified index agent and selective serotonin reuptake inhibitors (SSRIs) dominated among the duloxetine, TCAs, and venlafaxine cohorts. In the gabapentin and opioid cohorts, SSRIs and TCAs were the most frequently used among all the antidepressants. Duloxetine and SSRIs were the most used antidepressants in the pregabalin cohort. Across all cohorts (except for the opioid group), the most frequently used opioid was hydrocodone, followed by oxycodone. In the opioid cohort, tramadol topped the use of oxycodone (44.4 vs. 28.8%). More than 30% of study patients used NSAIDs, with most of the use among patients in the opioid cohort. Compared with the duloxetine cohort, the TCAs, venlafaxine, and gabapentin cohorts had significantly lower copayments for the index medication, while the pregabalin cohort had a significantly higher copayment (all p < 0.05). Approximately 8–10% of patients in each cohort were not on any anti-diabetic treatment during the 12-month pre-index period.
Table 3. Select medication use in the 12-month pre-index period.
Prior healthcare costs and utilization
Healthcare utilization and costs for the 12-month pre-index period are presented in . The inpatient hospitalization rate ranged between 31 and 42%, with a significantly higher admission rate among patients in the gabapentin, pregabalin, and opioids cohorts (all p < 0.05). A similar trend was observed for inpatient admissions specifically related to diabetes. Compared with the duloxetine cohort, the TCAs, venlafaxine, gabapentin, and opioid cohorts had a higher proportion of patients with DPNP-related admissions. The overall healthcare costs were between $29,313 and $39,265, with patients in the gabapentin and opioid cohorts costing the most (p < 0.05).
Table 4. Healthcare utilization and costs in the 12-month pre-index period.
Factors associated with duloxetine initiation vs. other DPNP medications
presents the results from the logistic regression analyses with two different sets of explanatory variables. In the models that focused on prior pain medication use (), we found that patients with any prior use of pregabalin (OR ranged 1.42–2.64, all p < 0.05) or duloxetine (OR ranged 1.67–4.33, all p < 0.05) were more likely to initiate duloxetine than any of the other DPNP therapies. Individuals with pre-index use of SSRI were more likely to begin treatment with duloxetine than TCAs (OR = 1.38, 95% CI = 1.08–1.76) or pregabalin (OR = 1.20, 95% CI = 1.01–1.42), and those with prior use of gabapentin were more likely to initiate duloxetine than TCAs (OR = 1.28, 95% CI = 1.03–1.59), venlafaxine (OR = 1.81, 95% CI = 1.33–2.46), or opioids (OR = 1.65, 95% CI = 1.40–1.94). Patients from other DPNP treatment cohorts, except for those in the pregabalin cohort, were more likely to re-initiate the same treatment than begin treatment with duloxetine (all p < 0.05). Those who had prior use of insulin and OADs were more likely to begin treatment with duloxetine than TCAs (OR = 1.63, 95% CI = 1.09–2.44) or opioids (OR = 1.39, 95% CI = 1.03–1.86).
Table 5a. Factors associated with duloxetine initiation vs. other DPNP treatments – prior medication use.
In the models focusing on prior comorbidities (), we found that patients with low back pain were more likely to initiate duloxetine than TCAs (OR = 1.33, 95% CI = 1.08–1.64) or venlafaxine (OR = 1.37, 95% CI = 1.06–1.77). Patients with infections related to diabetes were less likely to initiate duloxetine than venlafaxine (OR = 0.38, 95% CI = 0.18–0.79), gabapentin (OR = 0.54, 95% CI = 0.32–0.89), or opioids (OR = 0.47, 95% CI = 0.28–0.78). Patients with a history of anxiety were less likely to begin treatment with duloxetine than all other DPNP treatment agents (all p < 0.05), except for TCAs. Overall, demographic characteristics such as age and gender were not associated with patient initiation of duloxetine.
Table 5b. Factors associated with duloxetine initiation vs. other DPNP treatments – comorbidities.
Discussion
Using a large US administrative claims database, this study was the first to examine the factors associated with DPNP treatment among commercially insured working-age adults and their dependents. Patients initiating different pharmacologic therapies for the management of DPNP had similar demographics but different comorbidities and prior medication utilization profiles. Prior experience of medications for DPNP appeared to be influential in determining the choice of medication for initiation. Several comorbidities also played an important role in medication selection.
Our analysis found similarities in patient profiles among DPNP study cohorts. Specifically, patients in the duloxetine, TCAs, venlafaxine, and pregabalin cohorts had similar demographic characteristics, diabetes and pain-related conditions, and prior healthcare utilization and costs, while those in the gabapentin and opioid cohorts were similar. However, prior medication history appeared to be very different among the study cohorts. Not surprisingly, higher antidepressant use was observed among duloxetine, TCAs, and venlafaxine patients, whereas higher anticonvulsant use was found among patients who initiated gabapentin or pregabalin. Over two-thirds of patients in all cohorts had prior use of any opioids, with those in the pregabalin and opioid cohorts having the highest and those in the venlafaxine cohort the least. As found in previous studies, the most frequently used opioids were hydrocodone and oxycodone for all patients except those in the opioid cohort, where hydrocodone and tramadol were the most commonCitation4,Citation13. Prior use of NSAIDs was similar across all cohorts except for the opioid group, where the use of NSAIDs was higher.
Similar to previous findings among patients with diabetic neuropathyCitation13,Citation29, cardiovascular disease was the most prevalent diabetes-related comorbidity. This is not surprising as diabetes itself is a risk factor of cardiovascular diseaseCitation34. As suggested in the literatureCitation13, patients in our study cohorts also had high rates of other neuropathic or painful conditions such as low back pain and osteoarthritis.
In our study, patients who initiated gabapentin or opioids had higher all-cause and diabetes-related healthcare costs over the 12-month pre-index period than those who initiated duloxetine. Since healthcare costs are typically highly correlated with health status, our results suggest that patients initiated on gabapentin or opioids might have worse health status than those initiated on duloxetine. Future studies may need to focus in this direction.
Although NSAIDs are not recommended for the management of DPNPCitation10,Citation35, more than 30% of DPNP patients in our study used NSAIDs before the most recent initiation of therapy for DPNP. The use of NSAIDs might be under-estimated because our data did not capture the over-the-counter utilization. The use of NSAIDs observed here might be driven by other pain-related conditions rather than DPNP, because patients with mixed painful peripheral neuropathic disorders were more likely to use NSAIDsCitation36. Since there is little evidence to support the efficacy of NSAIDs in treating DPNPCitation10 and the use of NSAIDs may result in renal impairmentCitation1, further research is necessary to examine the use of NSAIDs in clinical practice for treating DPNP. It is worth noting that 8–10% of DPNP patients did not receive any anti-diabetic treatment during the pre-index period. The extent to which diabetes is controlled among these patients is not clear. Guidelines suggest that, while treating DPNP, providers should also strive to manage glucose levels in order to alleviate the painful symptomsCitation4.
Prior history with DPNP-related medications appeared to be associated with choice of treatment. Patients with prior use of gabapentin, TCAs, venlafaxine, or opioids were more likely to re-initiate the same treatment than to begin treatment with duloxetine. Furthermore, patients with pre-index use of duloxetine were more likely to re-initiate duloxetine therapy even though we required a 90-day gap in available therapy in the study design. However, pre-index use of pregabalin was consistently a significant factor associated with initiating duloxetine therapy across the models examined. This might be a result of the fact that duloxetine and pregabalin are the only two agents approved by the FDA for the management of DPNP. It is possible that providers who initially treated DPNP patients with pregabalin would likely switch them to duloxetine rather than other agents in the case of treatment failure.
Patients’ pre-existing comorbidities appeared to be associated with which DPNP medication patients used. Patients with a history of low back pain were more likely to initiate duloxetine than TCAs and venlafaxine, however these patients were more likely to initiate opioids rather than duloxetine therapy. Other pre-existing comorbidities associated with a decreased likelihood of duloxetine initiation were anxiety disorders and infections related to diabetes. Even though the current study could not examine the clinical justification for using one pharmacologic agent over another and how the comorbidities have impact on the medication selection, the empirical results from this study could serve as hypotheses-generating to assess areas where further exploration may be warranted.
Our study had several limitations. First, a selection bias could exist where DPNP patients had unobserved heterogeneity that correlated with levels of disease severities and choice of medications. Despite the best efforts to capture patient heterogeneity by baseline demographic and clinical characteristics, this retrospective study design was unable to control for unobserved confounding characteristics. In addition, we were unable to control for physician characteristics (e.g., gender, years in practice), prescribing behaviors and related preferences. We also lacked information regarding health plan benefit structure such as formulary and copayment structure. These provider and organizational characteristics may influence the access and choice of medications. Also, patients’ actual medical records were not available for this study and the identification of DPNP, diabetes and pain-related complications and comorbidities were based on the ICD-9-CM diagnosis codes recorded in the medical administrative claims. If such information was not correctly recorded, we may have misclassified patients in regards to their clinical conditions. Because patient chart reviews and physician notes were not captured in the data, we were unable to investigate why patients chose a specific medication or switched to different medications. Also, due to the administrative nature of our data source, we were unable to directly measure the severity of DPNP or the pain intensity, which may be associated with the choice of medications. Finally, causality of factors associated with the initiation of DPNP medications examined in the study could not be inferred because of the retrospective design. Because the study was conducted with a retrospective cohort design using a large US administrative claims database, findings from this research may need to be further confirmed with a prospective study where patients’ case notes rather than diagnosis codes are collected in order to better assess disease severity and minimize the impact of heterogeneity.
Conclusion
Among DPNP patients, those treated with duloxetine, TCAs, venlafaxine, or pregabalin had similar demographics, comorbidities, and pre-index healthcare costs, while patients who initiated gabapentin and opioids possessed similarity in these characteristics. Findings from this analysis suggest that a history of low back pain and prior use of DPNP-related medications were associated with duloxetine initiation when compared with other DPNP treatments.
Transparency
Declaration of funding
The funding of this project was provided by Eli Lilly and Company. The statements contained in this paper are solely those of the authors and no endorsement by Eli Lilly and Company should be inferred or implied.
Declaration of financial/other relationships
Z.Z. and Y.Z. are employed by Eli Lilly and Company and own company stock. S.C., N.W., L.B., and K.F. are employed by United BioSource Corporation, and no financial conflicts need to be declared.
Acknowledgments
No assistance in the preparation of this article is to be declared.
References
- Casellini CM, Vinik AI. Clinical manifestations and current treatment options for diabetic neuropathies. Endocr Pract 2007;13:550-66
- Cole BE. Diabetic peripheral neuropathic pain: recognition and management. Pain Med 2007;8:S27-32
- Boulton AJ. Management of diabetic peripheral neuropathy. Clin Diabetes 2005;23:9-15
- Boulton AJ, Vinik AI, Arezzo JC, et al. Diabetic neuropathies: a statement by the American Diabetes Association. Diabetes Care 2005;28:956-62
- Boulton AJ, Malik RA, Arezzo JC, et al. Diabetic somatic neuropathies. Diabetes Care 2004;27:1458-86
- Daousi C, MacFarlane IA, Woodward A, et al. Chronic painful peripheral neuropathy in an urban community: a controlled comparison of people with and without diabetes. Diabet Med 2004;21:976-82
- Ballantyne JC, Mao J. Opioid therapy for chronic pain. N Engl J Med 2003;349:1943-53
- Hall GC, Carroll D, Parry D, et al. Epidemiology and treatment of neuropathic pain: The UK primary care perspective. Pain 2006;122:156-62
- Galer BS, Gianas A, Jensen MP. Painful diabetic polyneuropathy: epidemiology, pain description, and quality of life. Diabetes Res Clin Pract 2000;47:123-8
- Argoff CE, Backonja MM, Belgrade MJ, et al. Consensus guidelines: treatment planning and options. Diabetic peripheral neuropathic pain. Mayo Clin Proc 2006;81:S12-25
- Gore M, Brandenburg NA, Dukes E, et al. Pain severity in diabetic peripheral neuropathy is associated with patient functioning, symptom levels of anxiety and depression, and sleep. J Pain Symptom Manage 2005;30:374-85
- Currie CJ, Poole CD, Woehl A, et al. The health-related utility and health-related quality of life of hospital-treated subjects with type 1 or type 2 diabetes with particular reference to differing severity of peripheral neuropathy. Diabetologia 2006;49:2272-80
- Gore M, Brandenburg NA, Hoffman DL, et al. Burden of illness in painful diabetic peripheral neuropathy: the patients' perspectives. J Pain 2006;7:892-900
- Jude EB, Jacob K. Managing painful neuropathy in diabetes. Pol Arch Med Wewn 2008;118:260-1
- Davies M, Brophy S, Williams R, et al. The prevalence, severity, and impact of painful diabetic peripheral neuropathy in type 2 diabetes. Diabetes Care 2006;29:1518-22
- Wu SC, Wrobel JS, Armstrong DG. Assessing the impact of pharmacologic intervention on the quality of life in diabetic peripheral neuropathic pain and fibromyalgia. Pain Med 2007;8:S33-42
- Jensen MP. Review of measures of neuropathic pain. Curr Pain Headache Rep 2006;10:159-66
- Goldstein DJ, Lu Y, Detke MJ, et al. Duloxetine vs. placebo in patients with painful diabetic neuropathy. Pain 2005;116:109-18
- Raskin J, Pritchett YL, Wang F, et al. A double-blind, randomized multicenter trial comparing duloxetine with placebo in the management of diabetic peripheral neuropathic pain. Pain Med 2005;6:346-56
- Gutierrez-Alvarez AM, Beltrán-Rodríguez J, Moreno CB. Antiepileptic drugs in treatment of pain caused by diabetic neuropathy. J Pain Symptom Manage 2007;34:201-8
- Wong MC, Chung JW, Wong TK. Effects of treatments for symptoms of painful diabetic neuropathy: systematic review. BMJ 2007;335:87
- Zin CS, Nissen LM, Smith MT, et al. An update on the pharmacological management of post-herpetic neuralgia and painful diabetic neuropathy. CNS Drugs 2008;22:417-42
- Várkonyi T, Kempler P. Diabetic neuropathy: new strategies for treatment. Diabetes Obes Metab 2008;10:99-108
- Robinson-Papp J, Simpson DM. Safety profile of treatment in diabetic peripheral neuropathic pain. Pain Med 2007;8:S43-9
- Chong MS, Hester J. Diabetic painful neuropathy: current and future treatment options. Drugs 2007;67:569-85
- Zhao Y, Ye W, Boye KS, et al. Prevalence of other diabetes-associated complications and comorbidities and its impact on health care charges among patients with diabetic neuropathy. J Diabetes Complications 2010;24:9-19
- Zhao Y, Ye W, Boye KS, et al. Healthcare charges and utilization associated with diabetic neuropathy: impact of type 1 diabetes and presence of other diabetes-related complications and comorbidities. Diabet Med 2009;26:61-9
- Boulanger L, Zhao Y, Foster TS, et al. Impact of comorbid depression or anxiety on patterns of treatment and economics outcomes among patients with diabetic peripheral neuropathic pain. Curr Med Res Opin 2009;25:1763-73
- Boulanger L, Zhao Y, Bao Y, et al. A retrospective study of factors associated with healthcare resource use and costs among patients with diabetic neuropathy. BMC Health Serv Res 2009;9:111
- Zhao Y, Sun P, Watson P. Medication adherence and healthcare costs among patients with diabetic peripheral neuropathic pain initiating duloxetine versus pregabalin. Curr Med Res Opin 2011;27:785-92
- Chen S, Wu N, Boulanger L, et al. The relationship between average daily dose, medication adherence, and healthcare costs among diabetic peripheral neuropathic pain patients initiated on duloxetine therapy. Pain Pract 2010;10:530-9
- Wu N, Chen S, Boulanger L, et al. The association of average daily dose of duloxetine on medication adherence and healthcare costs in patients with diabetic peripheral neuropathic pain. J Med Econ 2009;12:192-202
- Bureau of Labor Statistics. Consumer Price Index: January 2008
- Chaturvedi N. The burden of diabetes and its complications: trends and implications for intervention. Diabetes Res Clin Pract 2007;76:S3-12
- Barrett AM, Lucero MA, Le T, et al. Epidemiology, public health burden, and treatment of diabetic peripheral neuropathic pain: a review. Pain Med 2007;8:S50-62
- Gore M, Dukes E, Rowbotham DJ, et al. Clinical characteristics and pain management among patients with painful peripheral neuropathic disorders in general practice settings. Eur J Pain 2007;11:652-64
Appendix
Table A1. Diabetes- and pain-related comorbidities.