Cannabis Use in Patients with Gastroparesis

Abstract

Aim:

The primary aim was to determine the prevalence of cannabis use among patients hospitalized for gastroparesis. The secondary aim was to identify independent variables associated with cannabis use compared with noncannabis-related gastroparesis hospitalization.

Methods:

We use the nationwide inpatient sample database from January 2012 to December 2014. The patients included in this study were the ones with primary diagnosis of gastroparesis and cannabis use. The analysis was performed using the Statistical Package for the Social Sciences 27 (SPSS) and a multivariable regression was conducted to identify independent variables.

Results:

We found 50,170 patients with a primary diagnosis of gastroparesis. The prevalence of cannabis use among patients hospitalized for gastroparesis was 4.2%. Multivariate regression analysis was performed, adjusting for confounders. The variables found to increase the odds of cannabis use in gastroparesis populations independently were age interval of 18–35 and 36–50 years, male, Black and Asian, median household income 1–25th percentile, Medicaid insurance, no charge hospitalization, and smoking. Cannabis use was associated with lower odds of vomiting.

Conclusion:

Patients who used cannabis were younger and of African American, Asian, or Pacific Islander descent. They had Medicaid insurance and a lower median household income.

Introduction

Gastroparesis is a complex syndrome with cardinal symptoms of nausea, vomiting, early satiety, belching, bloating, upper abdominal pain, and delayed gastric emptying of solids in the absence of mechanical obstruction.¹ The common causes of gastroparesis are idiopathic, diabetes mellitus, medication induced, or postsurgical.²

The predominant symptoms of gastroparesis vary based on the underlying etiology. It can present with nausea (93%), vomiting (68–84%), abdominal pain (46–90%), early satiety (60–86%), postprandial fullness, and weight loss.^3,4 In a retrospective study, patients with idiopathic gastroparesis predominantly reported early satiety, postprandial fullness, and abdominal pain, whereas diabetic gastroparesis patients had more severe retching and vomiting.⁵ Severe abdominal pain was associated with anxiety, somatization, impaired quality of life, opiate use,⁶ and cannabis use.⁷

Pharmacotherapy in patients with gastroparesis is not very useful, with less than one-third of patients showing improvement in the symptoms after a year or longer on medical treatment.⁸ There are surgical options for medically refractory gastroparesis like gastric electrical stimulation (GES), pyloric intervention, and so on. Since GES has been introduced, clinical trials are being conducted for its efficacy. Some open-labeled studies have shown 1-year clinical response rates vary from 45% to 74%.^9–12 There has been newer promising pyloric-directed therapy, but randomized controlled trials need to be conducted to prove its clinical benefits.

Synthetic cannabinoids have been approved by the U.S. Food and Drug Administration for oral administration to treat anorexia associated with weight loss in AIDS patients and nausea and vomiting associated with cancer chemotherapy refractory to conventional antiemetic therapy.^13,14 Cannabis may be used in patients with gastroparesis for its antiemetic and orexigenic effects. In a retrospective study that involved 506 gastroparesis patients, 12% used cannabis and found it beneficial for severe nausea and abdominal pain.¹⁵

To our knowledge, there have been no clinical studies evaluating the use of cannabis in gastroparesis patients at the national level. In this study, we aimed to estimate the prevalence of cannabis use among patients hospitalized for gastroparesis and evaluate outcomes, risk factors, and trends compared with noncannabis-related gastroparesis hospitalizations.

Patients and Methods

Data source

The Nationwide Inpatient Sample (NIS) database from January 2012 to December 2014 was used for this study. The NIS database is developed for the Healthcare Cost and Utilization Project (HCUP) for decision making at the national, state, and community levels. It is the largest publicly available database designed to produce U.S. regional and national estimates of in-hospital health care access, utilization, charges, and outcomes. Unweighted, it contains more than seven million hospital stays each year. Weighted, it has >35 million hospital visits nationally.¹⁶ The NIS-HCUP database is weighted to reliably create discharge-level estimates, such as estimates of the total number of discharges with a diagnosis of gastroparesis in the United States.¹⁶

A weight was then calculated for each stratum, by dividing the number of universe discharges in that stratum, obtained from HCUP and the American Hospital Association data, by the number of NIS discharges in the stratum. Weighted estimates can be calculated by applying the discharge weights to the sample discharges. The NIS is a discharge-level database containing de-identified clinical and nonclinical data elements of each hospital stay. It includes information on patient's demographics (age, sex, race, median household income for zip code), principal diagnosis, and up to 29 secondary diagnoses using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) coding system as well as up to 15 procedures using the ICD-9-CM procedural codes. Hospital level data includes hospital location, teaching status, number of beds, and so on. It also provides information about hospital length of stay, hospital charges, and discharge disposition.

Study population

Patients with the primary diagnosis of gastroparesis (ICD-9 code 536.3) were included in the study. Patients with concomitant use of cannabis were identified. The ICD-9-CM diagnostic codes used to identify cases of cannabis use are given in Table 1. The comorbid conditions were identified in the gastroparesis population using ICD-9 codes: nausea (787.02), vomiting (787.03), early satiety (780.94), bloating\flatulence (787.3), and epigastric abdominal pain (789.06).

Table 1.

International Classification of Diseases, 9th Revision, Clinical Modification Codes Study Outcomes and Variables

Cannabis use disorder
304.30	Unspecified cannabis dependence
304.31	Continuous cannabis dependence
304.32	Episodic cannabis dependence
305.20	Unspecified cannabis abuse
305.21	Continuous cannabis abuse
305.22	Episodic cannabis abuse

The primary outcome was to determine the prevalence of cannabis use among patients hospitalized for gastroparesis and the prevalence of cannabis-related gastroparesis hospitalization. The secondary outcome was to access health care utilization that included length of stay and hospital charges and independent predictors associated with cannabis- related gastroparesis hospitalization. Multiple confounders were identified and accounted for in the analysis like age, sex, race, median household income, payer source, and comorbid conditions. The hospitalization cost was adjusted for inflation using the consumer price index. The cost was presented in 2014 U.S. dollars.

Study analysis

All analysis was performed using the Statistical Package for the Social Sciences 27 (SPSS). Data were weighted using the discharge-level weight variable for estimating national U.S. numbers. Continuous variables like age were reported as mean and standard deviation. Categorical factors were compared with chi-square tests. Values of p<0.05 indicate a statistically significant association. Univariate analysis was performed to compare patients' demographics, morbidity, and hospitalization. Weighted multivariable regression was conducted to identify independent variables associated with cannabis-related gastroparesis hospitalization. Because the NIS-HCUP database does not contain any patient identifiers, institutional review board approval was not required.

Results

A total of 50,170 hospitalizations from 2012 to 2014 with a primary diagnosis of gastroparesis were included in the study. Of these, 2100 (4.2%) gastroparesis hospitalization had cannabis use.

The gastroparesis patients with cannabis use were 12 years younger compared with the control group. The age interval of 18–35 years had highest number of cannabis users, 1155 (55%), with odds ratio (OR)=4.25; 95% confidence interval (CI), 3.89–4.64 (p<0.0001, Table 2).

Table 2.

Demographic Factors of Cases with Gastroparesis and Cannabis Use

	Cannabis use, n (%)	No cannabis use, n (%)	p
Age	35.75 SD 10.89	47.62 SD 18.36	<0.0001
Age interval, years
0–17	30 (1.4)	2070 (4.3)
18–35	1155 (55)	10,740 (22.3)
36–50	665 (31.7)	14,650 (30.5)
51–65	250 (11.9)	12,035 (25)
>65	0	8575 (17.8)
Sex
Male	780 (37.1)	12,580 (26.2)	<0.0001
Female	1320 (62.9)	35,490 (73.8)
Race/ethnicity
White	1145 (54.5)	30,435 (63.3)	<0.0001
Black	625 (29.8)	9655 (20.1)
Hispanic	165 (7.9)	4280 (8.9)
Asian or Pacific	25 (1.2)	385 (0.8)
Islander
Native American	10 (0.5)	240 (0.5)
Other	45 (2.1)	1070 (2.2)
Median zip code income quartile
0–25%	825 (40.1)	14,995 (31.8)	<0.0001
26–50%	580 (28.2)	12,425 (26.3)
51–75%	415 (20.2)	10,980 (23.3)
76–100%	235 (11.4)	8790 (18.6)
	Missing=45	Missing=880
Primary payer			<0.0001
Medicare	365 (17.5)	19,205 (40)
Medicaid	750 (35.9)	9595 (20)
Private insurance	510 (24.4)	15,195 (31.6)
Self-pay	335 (16)	2475 (5.2)
No charge	50 (2.4)	275 (0.6)
Other	80 (3.8)	1270 (2.6)
	Missing=10	Missing=55

SD, standard deviation.

Comparing the cannabis cohort to noncannabis cohort, men was in higher number in cannabis group, 37.1% versus 26.2%; Black were in higher number with OR=1.69, 95% CI, 1.53–1.86, p<0.0001; 8.3% more patients in the median household income 0–25th percentile with OR=1.43, 95% CI, 1.31–1.56, p<0.0001; and 7.2% less patients in the cannabis cohort with OR=0.56, 95% CI, 0.50–0.65, p<0.0001. Medicaid and no charge patients had higher odds of being in the cannabis cohort with OR=2.23, 95% CI, 2.03–2.44, p<0.0001, and OR=4.24, 95% CI, 3.13–5.75, p<0.0001, respectively.

From 2012 to 2014, the prevalence of cannabis use in gastroparesis patients increased by 38.74%, from 555 to 770. The percentage of gastroparesis patients with cannabis use increased between 2012 (3.2%) and 2014 (4.7%; Table 3).

Table 3.

Yearly Trends of Gastroparesis Admissions With and Without Cannabis Use

Year	Total gastroparesis hospitalization	Gastroparesis hospitalization with cannabis use	Percentage of gastroparesis with cannabis use (%)
2012	17,220	555	3.2
2013	16,460	775	4.7
2014	16,490	770	4.7

The mean length of stay was shorter in the cannabis cohort than of the noncannabis cohort, 3.54 days versus 5.07 days (p<0.0001). The mean total hospitalization costs were lower in gastroparesis patients with cannabis use than noncannabis patients, $24,048.23 versus $35,512.58 (p<0.0001). Similarly, the cannabis group had fewer chronic conditions than those without cannabis (mean of 4.96 and 5.13, respectively, p<0.0001). West region hospitals had 2.9% higher cannabis users than noncannabis users (Table 4).

Table 4.

Comparison of Hospital Costs and Characteristics Between Cannabis and Noncannabis Cohort

	Cannabis, n (%)	Noncannabis, n (%)	p
Length of stay	3.54 SD 3.13	5.07 SD 5.17	<0.0001
Number of chronic conditions	4.96 SD 2.4	5.13 SD 3.00	0.02
Total	24,048.23 SD	35,512.58 SD	<0.0001
Charges in dollars	22,172.76	44,697.16
Hospital region
Northeast	325 (15.5)	8035 (16.7)	0.003
Midwest	405 (19.3)	9405 (19.6)
South	985 (46.9)	23,230 (48.3)
West	385 (18.3)	7400 (15.4)
Teaching status of the hospital			0.07
Rural hospital	530 (25.2)	12,375 (25.7)
Urban nonteaching	860 (41)	18,515 (38.5)
Urban teaching	710 (33.8)	17,180 (35.7)

Gastroparesis patients with cannabis use had lower odds of having nausea, vomiting, early satiety, and bloating/flatulence. Epigastric abdominal pain, smoking, and alcohol had higher odds in the cannabis cohort than those without cannabis use (Table 5).

Table 5.

Comparison of Morbidity Between Cannabis Users and Those Without Cannabis Use

Factor	Cannabis, n (%)	Noncannabis, n (%)	OR	p
Nausea	0	410 (0.9)	0.991 (0.991–0.992)	<0.0001
Vomiting	10 (0.5)	530 (1.1)	0.43 (0.23–0.80)	0.006
Early satiety	0	215 (0.4)	0.996 (0.995–0.996)	0.002
Bloating/flatulence	0	225 (0.5)	0.995 (0.995–0.996)	0.002
Epigastric abdominal pain	60 (2.9)	990 (2.1)	1.40 (1.10–1.82)	0.012
Diabetes mellitus	235 (11.2)	9465 (19.7)	0.51 (0.49–0.59)	<0.0001
Smoking	915 (43.6)	7290 (15.2)	4.32 (3.95–4.73)	<0.0001
Alcohol	20 (1)	255 (0.5)	1.80 (1.14–2.85)	0.01

OR, odds ratio.

Multivariate regression analysis was performed adjusting for confounders. The variables found to increase the odds of cannabis use in gastroparesis populations independently were in those in the age interval of 18–35 and 36–50 years, sex male, Black and Asian, median household income 1–25th percentile, Medicaid insurance, no charge hospitalization, and smoking. Cannabis use was associated with lower odds of vomiting. Patients who were using cannabis and were admitted to hospital for gastroparesis had low odds of suffering from diabetes mellitus. Table 6 provides the independent predictors of cannabis use in gastroparesis patients.

Table 6.

Multivariable Logistic Regression Analysis of Factors Associated with Cannabis Use in Patients Admitted for Gastroparesis

Factor	OR	95% CI	p
Age, years
0–17	R 5.15
18–35	2.36	3.55–7.48	0.0001
36–50	1.23	1.62–3.44	0.0001
51–65	0.000	0.83–1.82	0.294
>65		0.00–0.00	0.969
Sex
Male	1.75	1.59–1.92	0.0001
Race
White	R
Black	1.45	1.30–1.61	0.0001
Hispanic	0.92	0.77–1.10	0.361
Asian or Pacific	2.20	1.42–3.39	0.0001
Islander
Native American	0.85	0.44–1.63	0.621
Other	1.3	0.95–1.79	0.1
Median zip code income quartile 0–25%
26–50%	1.45	1.25–1.68	0.0001
51–75%	1.37	1.18–1.60	0.0001
76–100%	1.20	1.02–1.40	0.03
	R
Primary payer
Medicare	R 1.87
Medicaid	1.10	1.64–2.13	0.0001
Private insurance	2.61	0.95–1.27	0.196
Self-pay	3.47	2.22–3.07	0.0001
No charge	1.71	2.48–4.85	0.0001
Other		1.32–2.21	0.0001
Nausea	0.000	0.00	0.99
Vomiting	0.52	0.27–0.98	0.04
Early satiety	0.00	0.000	0.99
Bloating/flatulence	0.00	0.00	0.99
Epigastric abdominal pain	1.10	0.82–1.44	0.54
Diabetes mellitus	0.70	0.60–0.81	0.0001
Smoking	3.15	2.86–3.46	0.0001
Alcohol	1.25	0.78–2.01	0.36

CI, confidence interval; R, reference value.

Discussion

In this study, we assessed the prevalence of cannabis use among patients hospitalized for gastroparesis. It was shown that 4.2% of patients diagnosed with gastroparesis use cannabis. Patients using cannabis were younger; were African American, Asian, or Pacific Islander; had a lower median household income; and had Medicaid insurance. Patients using cannabis had shorter hospital stays, which led to lower hospital costs. Patients using cannabis had decreased odds of experiencing nausea, vomiting, early satiety, bloating, and abdominal pain. Still, after adjusting for demographics and socioeconomic factors, they only had decreased odds of vomiting.

In our study, the prevalence of cannabis use among patients hospitalized for gastroparesis was 4.2%. The reported use in a gastroparesis cohort study was lower than in the general U.S. population: 13.9% in those 12 years and older, according to a 2016 survey.¹⁷ The prevalence of cannabis use may be owing to different periods. There is a 16% increase in the prevalence of cannabis use since 2006¹⁸; second, patients might not report cannabis use because of fear of repercussions.⁷

In this analysis, cannabis use varied with age. The highest prevalence was among adults 18–35 years old, and the lowest prevalence was among those 65 years and older. This study is in agreement with a 2016 national survey on drug use and health, which showed the highest prevalence among young adults aged 18–25 years (33%), and the lowest prevalence among those aged 65 years or older (3.3%).¹⁷ Like the U.S. general population, men were more likely to be a cannabis user.¹⁷ Cannabis use was more prevalent in African Americans or the Asians/Pacific Islanders. In a 2016 national survey, cannabis use was more prevalent in African Americans (13.6%), intermediate in the Pacific Islander (8.6%), but less prevalent in Asians.¹⁷ Low cannabis prevalence in Asians contrasts with the results of this study, which showed higher cannabis prevalence in Asian/Pacific Islanders. The possible explanation for this is that NIS-HCUP reported Asian and Pacific Islanders as a combined race.

The results showed that adults within the 1–25th percentile of median household income had a higher rate of cannabis use, at 1.45 times, than adults within the income range in the 76–100th percentile. According to a 2015 U.S. epidemiology study, adults with an income of less than $20,000 annually had a higher rate of cannabis use, at 2.5 times, than adults with an income of at least $70,000 annually.¹⁸

In this study, cannabis users had decreased odds of experiencing nausea, vomiting, early satiety, bloating, and epigastric abdominal pain than those who did not use cannabis.

However, after adjusting for confounders on multivariate analysis, only vomiting was associated with decreased odds that were statistically significant. Do these patients' using cannabis have lower severity of symptoms, or is the decrease in vomiting a consequence of cannabis use? Unfortunately, this study does not provide a definitive answer to this question. According to a study published involving 24 gastroparesis patients, cannabis use was associated with significant improvement in the Gastroparesis Cardinal Symptoms Index total symptoms, and abdominal pain score with a minimum of 60-day use. However, that study had drawbacks owing to the small sample size, lack of randomization, and blinded methodology.¹⁹ In another study involving 506 gastroparesis patients, cannabis users had higher nausea/vomiting and upper abdominal pain subscores⁷; 81% patients using cannabis rated their benefits from cannabis as better or much better.⁷

According to our study, cannabis use was associated with shorter lengths of hospital stays and the health care system's overall cost. A possible reason could be that cannabis users were mainly younger, which means they had less chronic conditions, as chronic conditions increase with age.²⁰ They had lower vomiting rates that would lead to an early start of an oral diet and discharge compared with those not using cannabis. An observational study involving 806 postoperative patients depicted that nausea and vomiting prolonged the hospital stay.²¹

This study has several limitations. We were unable to obtain information on the severity of symptoms (for nausea, vomiting, etc.) for hospitalization. This database uses ICD-9 codes, which are subject to inaccurately entered or missing codes.²² Owing to NIS database limitations, it did not allow us to identify gastroparesis' etiology (idiopathic, diabetes, etc.) and information about gastric emptying study. It involves in-hospital patients, so we evaluated the prevalence of cannabis use with outpatient gastroparesis patients. Finally, our study was observational, so randomization was not performed, but we performed multivariate analysis to adjust for confounders.

To conclude, the prevalence of cannabis use among patients hospitalized for gastroparesis was 4.2%. Patients who used cannabis were younger and of African American, Asian, or Pacific Islander descent. They had Medicaid insurance and a lower median household income. This study showed cannabis use in gastroparesis patients was associated with lower odds of vomiting. It is not clear whether cannabis use decreases the vomiting or if they had a lower rate of vomiting severity before the use of cannabis. Randomized, controlled, or prospective clinical trials are needed to address it.

Footnotes

Authors' Contributions

S.S., Z.I.T., M.A., and M.A.A.: conception and design of the work; S.S.: analysis and interpretation of data; S.S., Z.I.T., M.A., and M.A.A.: drafting of the article; A.T. and T.L.A.: final approval of the article. All authors approved the final version of the article to be published.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Abbreviations Used

References

Camilleri

, Parkman

, Shafi

, et al. Clinical guideline: Management of gastroparesis. Am J Gastroenterol, 2013; 108(1):18–37; quiz 3837 quiz 38; doi: 10.1038/ajg.2012.373.

Camilleri

, Chedid

, Ford

, et al. Gastroparesis. Nat Rev Dis Primers, 2018; 4(1):41:2018; doi: 10.1038/s41572-018-0038-z.

Soykan

, Sivri

, Sarosiek

, et al. Demography, clinical characteristics, psychological and abuse profiles, treatment, and long-term follow-up of patients with gastroparesis. Dig Dis Sci, 1998; 43(11):2398; doi: 10.1023/a:1026665728213.

Hoogerwerf

, Pasricha

, Kalloo

, et al. Pain: The overlooked symptom in gastroparesis. Am J Gastroenterol, 1999; 94(4):1029; doi: 10.1111/j.1572-0241.1999.01008.x.

Parkman

, Yates

, Hasler

, et al. Similarities and differences between diabetic and idiopathic gastroparesis. Clin Gastroenterol Hepatol, 2011; 9(12):1056; quiz e133; doi: 10.1016/j.cgh.2011.08.013.

Parkman

, Wilson

, Hasler

, et al. Abdominal pain in patients with gastroparesis: Associations with gastroparesis symptoms, etiology of gastroparesis, gastric emptying, somatization, and quality of life. Dig Dis Sci, 2019; 64(8):2242; doi: 10.1007/s10620-019-05522-9.

Parkman

, Sharkey

, Nguyen

, et al. Marijuana use in patients with symptoms of gastroparesis: Prevalence, patient characteristics, and perceived benefit. Dig Dis Sci, 2020; 65(8):2311; doi: 10.1007/s10620-019-05963-2.

Pasricha

, Yates

, Nguyen

, et al. Outcomes and factors associated with reduced symptoms in patients with gastroparesis. Gastroenterology, 2015; 149(7): 1762–1774.e41774e4; doi: 10.1053/j.gastro.2015.08.008.

Heckert

, Sankineni

, Hughes

, et al. Gastric electric stimulation for refractory gastroparesis: A prospective analysis of 151 patients at a single center. Dig Dis Sci, 2016; 61(1):168–175; doi: 10.1007/s10620-015-3837-z.

10.

McCallum

, Lin

, Forster

, et al. Gastric electrical stimulation improves outcomes of patients with gastroparesis for up to 10 years. Clin Gastroenterol Hepatol, 2011; 9(4):314–319.e1319e1; doi: 10.1016/j.cgh.2010.12.013.

11.

Velanovich

. Quality of life and symptomatic response to gastric neurostimulation for gastroparesis. J Gastrointest Surg, 2008; 12(10): 1656–1662; discussion 1662; doi: 10.1007/s11605-008-0655-z.

12.

McCallum

, Snape

, Brody

, et al. Gastric electrical stimulation with Enterra therapy improves symptoms from diabetic gastroparesis in a prospective study. Clin Gastroenterol Hepatol, 2010; 8(11): 947–954; quiz e116; doi: 10.1016/j.cgh.2010.05.020.

13.

Beal

, Olson

, Laubenstein

, et al. Dronabinol as a treatment for anorexia associated with weight loss in patients with AIDS. J Pain Symptom Manag, 1995; 10(2):89–97; doi: 10.1016/0885-3924(94)00117-4.

14.

May

, Glode

. Dronabinol for chemotherapy-induced nausea and vomiting unresponsive to antiemetics. Cancer Manag Res, 2016; 8:49–55; doi: 10.2147/CMAR.S81425.

15.

Parkman

, Sharkey

, Nguyen

, et al. Marijuana use in patients with symptoms of gastroparesis: Prevalence, patient characteristics, and perceived benefit. Dig Dis Sci, 2020; 65(8):2311–2320; doi: 10.1007/s10620-019-05963-2.

16.

HCUP National Inpatient Sample (NIS). Healthcare Cost and Utilization Project (HCUP). 2012. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/nisoverview.jsp

17.

Center for Behavioral Health Statistics and Quality. 2016. National Survey on Drug Use and Health: Detailed Tables. Substance Abuse and Mental Health Services Administration; Health and Human Services: Rockville, MD; 2017.

18.

Hasin

, Saha

, Kerridge

, et al. Prevalence of Marijuana Use Disorders in the United States Between 2001–2002 and 2012–2013. JAMA Psychiatry, 2015; 72:1235; doi: 10.1001/jamapsychiatry.2015.1858

19.

Barbash

, Mehta

, Siddiqui

, et al. Impact of cannabinoids on symptoms of refractory gastroparesis: A single-center experience. Cureus 2019;11(12):e6430;, 2019; doi: 10.7759/cureus.6430.

20.

Bielefeldt

. Regional differences in healthcare delivery for gastroparesis. Dig Dis Sci, 2013; 58(10):2789–2798; doi: 10.1007/s10620-013-2643-8.

21.

Mc Loughlin

, Terrasa

, Ljungqvist

, et al. Nausea and vomiting in a colorectal ERAS program: Impact on nutritional recovery and the length of hospital stay. Clin Nutr ESPEN, 2019; 34:73–80; doi: 10.1016/j.clnesp.2019.08.010.

22.

Klabunde

, Warren

, Legler

. Assessing comorbidity using claims data: An overview. Med Care, 2002; 40(8 Suppl.):IV-26–IV-35; doi: 10.1097/00005650-200208001-00004.