Identifying vulnerable populations with symptomatic cholelithiasis at risk for increased health care utilization : Journal of Trauma and Acute Care Surgery

Secondary Logo

Journal Logo

INDEPENDENT SUBMISSIONS

Identifying vulnerable populations with symptomatic cholelithiasis at risk for increased health care utilization

Shenoy, Rivfka MD, MS; Kirkland, Patrick MD; Jackson, Nicholas PhD, MPH; DeVirgilio, Michael MD; Zingmond, David MD, PhD; Russell, Marcia M. MD; Maggard-Gibbons, Melinda MD, MSHS

Author Information
Journal of Trauma and Acute Care Surgery: December 2022 - Volume 93 - Issue 6 - p 863-871
doi: 10.1097/TA.0000000000003778
  • Free
  • SDC
  • Infographic
  • Best Of

An estimated 15% of Americans have gallstones and cost of their treatment surpasses six billion dollars annually.1–3 Symptomatic cholelithiasis (SC) occurs in 10% of patients with gallstones within 5 years and can progress to more complex manifestations, such as acute cholecystitis, choledocholithiasis, or gallstone pancreatitis.2,4–7

Gallstones are a leading cause for emergency department (ED) visits1,2 and patients may be first diagnosed with SC in the ED. If urgent intervention is not performed, patients are typically discharged with a referral to a surgeon or primary care physician (PCP). However, some will experience repeat ED visits (ED revisits) for biliary-related symptoms and progression of disease while waiting for this follow-up appointment.8,9

There is limited research on which patients with SC are at risk for this increased ED utilization.10–12 It is also unknown what system and patient factors are associated with increased ED revisits for this common disorder. Such an understanding may guide implementations to improve equality in the treatments of SC. The aims in this study are to (1) assess longitudinal ED utilization and subsequent surgery for patients presenting with SC; and (2) identify patient, geographic, and hospital characteristics associated with ED revisits following index diagnosis of SC.

METHODS

Data Sources and Ethics Review

The primary data source is the California Office of Statewide Health Planning and Development (OSHPD) 2016 to 2018.13 The three databases analyzed were (1) ED database (EDD), (2) patient discharge database (PDD) which captures acute care hospitals and (3) ambulatory surgery database, which captures ambulatory surgery care centers. These were merged using unique patient record linkage numbers and were linked to: US Census Bureau data (socioeconomic and acculturation data at zip code tract area [ZCTA] level),14 American Medical Association Physician Masterfile for 2017 (combined with US Census data to create a county level PCP to total county population ratio)15 and the OSHPD hospital annual financial disclosure reports (desk-audited data collected from all acute care hospitals).16 This study was approved by the OSHPD Committee of the Protection of Human Subjects. This study conforms with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines and a complete checklist has been uploaded as Supplemental Digital Content (Supplemental Content A, https://links.lww.com/TA/C681).

Patient Cohort and Follow-Up

We identified patients with a diagnosis of SC between 7/1/2016 and 12/31/2017 who were discharged from the ED (Supplemental Content B, https://links.lww.com/TA/C682, for primary ICD-10 diagnosis codes). Patients were excluded if they had an ED visit for SC within the prior 6 months, were admitted at the index ED visit, received cholecystectomy at index ED visit, or received a cholecystostomy tube within 1 year (not eligible for standard follow-up).

Future biliary-related visits (including ED visits or hospital admission) recorded in the EDD and PDD, and receipt of cholecystectomy (see Supplemental Content C for codes, https://links.lww.com/TA/C683) recorded in the ambulatory surgery database or PDD were tracked for 1 year after index visit.

Outcome of Interest Definition

The main outcome of interest was ED revisits (defined as zero vs one or more) with a biliary-related primary diagnosis. For patients with ED revisits, the variable was stratified as high utilizers (≥2 visits) or low utilizers (1 visit).

Variable Definition and Selection

Two main regressors of interest were patient race/ethnicity and insurance status obtained from the OSHPD administrative database. Race/ethnicity was operationalized by combining the race and ethnicity variables. Categories included non-Hispanic White, Hispanic, non-Hispanic Black or African-American, non-Hispanic Asian and other/unknown/not reported. Non-Hispanic American Indian or Alaska Native and non-Hispanic Native Hawaiian or Pacific Islander identifying patients were collapsed into the other category. Insurance status was operationalized into five categories: Medicare, Medicaid, private, self-pay, and other. Patient characteristics included race/ethnicity, insurance, age (years, operationalized as continuous in tables and text, analyses repeated as a categorical variable and mentioned in text as a subgroup analysis), sex (male/female). Full list of geographic and hospital variables is shown in Supplemental Content D, https://links.lww.com/TA/C684, along with rationale for excluding other considered variables.17,18

Statistical Analysis

There were minimal missing data with the following exceptions: race/ethnicity (0.8%), bilingual zip-code (1.0%), zip-code median income (1.1%), hospital characteristics (0.4%). Data were not imputed, and the results provided are for patients with complete variables.

Descriptive data included mean/median with standard deviation [SD]/interquartile region [IQR] for continuous variables and frequencies/percentages for categorical variables. To determine the appropriateness of modeling continuous variables linearly, Lowess plots were used to evaluate the functional form of the variable versus the primary outcome of interest. Two-group comparisons with ED revisits (yes/no) and ED utilization (high/low) were done using Wilcoxon Rank-Sum tests for continuous variables and χ2 tests or Fisher's exact test for categorical variables. Variables were chosen for inclusion in multivariable models if they had a univariate p value less than 0.2, with the exception of the main regressors. As such, not all multivariable models included the same regressors. With multivariable logistic regression models, we estimated the association of race/ethnicity and insurance status with each of the outcomes of interest, adjusting for the remaining covariates, and clustering standard errors by hospital of initial presentation. For the ED revisits outcome, we specified the following models: Model 1 included only patient variables; Model 2 added geographic variables; Model 3 added hospital variables. The same models were used for the ED utilization outcome analyses, with the exception of the R and S variables in Model 3. All analyses were performed using STATA v16.0 (StataCorp, College Station, TX), using two-sided tests and an α of 0.05.

RESULTS

Patient Characteristics

We identified 34,427 patients who presented to the ED with SC between July 1, 2016, and December 31, 2017. The majority were Hispanic (50.8%) and female (72.8%) (Table 1). Median age was 40 years (IQR, 29–56 years). The largest proportion of patients were privately insured (46.1%) followed by Medicaid insured (39.7%); 5.0% were self-pay (5.0%). Approximately one-third came from a bilingual zip code and approximately two-thirds lived in a county that had a PCP to population ratio below the state median. Patients who initially presented to a safety-net hospital made up 23.8% of the population and 71.9% of patients presented to a nonprofit hospital.

TABLE 1 - Patient and Hospital Characteristics for OSHPD Population With Symptomatic Gallstones by Emergency Department Revisits
Recurrent ED Use
Variables Total No Recurrent ED Use Recurrent ED Use p
34,427 (100%) 27,946 (81.2%) 6,481 (18.8%)
Patient characteristics
Race/ethnicity, n (%)* < 0.001
 White 11,516 (33.5%) 9,655 (83.8%) 1,861 (16.2%)
 Hispanic 17,475 (50.8%) 13,772 (78.84%) 3,703 (21.2%)
 Black or African American 2,147 (6.2%) 1,663 (77.5%) 484 (22.5%)
 Asian 1,904 (5.5%) 1,686 (88.6%) 218 (11.4%)
 Other 1,385 (4.0%) 1,170 (84.5%) 215 (15.5%)
Insurance, n (%)* < 0.001
 Private 15,883 (46.1%) 13,642 (85.9%) 2,241 (14.1%)
 Medicare 2,900 (8.4%) 2,551 (88.0%) 349 (12.0%)
 Medicaid 13,662 (39.7%) 10,189 (74.6%) 3,473 (25.4%)
 Self-pay 1,713 (5.0%) 1,349 (78.7%) 364 (21.3%)
 Other 269 (0.8%) 215 (79.9%) 54 (20.1%)
Age, median [IQR] 40 [29, 56] 42 [30, 58] 33 [25, 48] < 0.001
Sex (female), n (%)** 25,071 (72.8%) 20,144 (72.1%) 4,927 (76.0%) < 0.001
Geographic characteristics
Zip-code income, median [IQR] 63,748 [$50,404–$82,128] $64,164 [$50,903–$83,095] $59,776 [$47,037–$74,050] < 0.001
Bilingual zip-code, n (%)** 11,955 (34.7%) 9,376 (33.6%) 2,579 (39.8%) < 0.001
PCP ratio§ below median, n (%)** 22,562 (65.5%) 17,925 (64.1%) 4,637 (71.6%) < 0.001
Hospital Characteristics
Hospital control, n (%)* < 0.001
 Investor 5,656 (16.4%) 4,484 (79.3%) 1,172 (20.7%)
 County/city 1,870 (5.4%) 1,467 (78.4%) 403 (21.6%)
 Nonprofit 24,755 (71.9%) 20,322 (82.1%) 4,433 (17.9%)
 District 2,146 (6.3%) 1,673 (78.0%) 473 (22.0%)
Hospital size (licensed beds), median [IQR] 302 [199, 434] 302 [202, 435] 291 [193, 420] < 0.01
Safety-net status (yes), n (%)** 8,209 (23.8%) 6,260 (22.4%) 1,949 (30.1%) < 0.001
Teaching hospital (yes), n (%)** 14,746 (42.8%) 12,071 (43.2%) 2,675 (41.3%) < 0.01
ED volume (total annual visits), median [IQR] 69,726 [46,395–94,002] 69,726 [46,395–94,002] 69,726 [45,827–89,791] 0.07
*Percentages for Race/Ethnicity, Insurance, and Hospital control represent row percentages (e.g., 84.4% of non-Hispanic White patients had no ED revisits out of all non-Hispanic White patients) with denominators shown on the same row in “Total” column (percentages in the Total column reflect column percentages).
**Percentages for Sex, PCP Ratio, Safety-Net Status, and Teaching Hospital Status represent column percentages (e.g., 72.0% of all patients who did not have ED revisits were female) with denominators shown at the top of each column.
Median income in the patient's zip code.
Bilingual status defined as >50% of zip-code speaking a language other than English.
§Ratio of number of PCPs to total population in the patient's county.

One-Year Follow-Up

At 1 year, 18.8% patients had an ED revisit for a biliary-related diagnosis (n = 6,481) (Fig. 1). Over two-thirds (72.4%) of these patients returned once to the ED for a biliary-related problem, while 27.6% returned two or more times. A subgroup exploration of the 4,693 patients who returned once to the ED for a biliary-related problem revealed that almost two-thirds came back for symptomatic cholelithiasis (calculus of gallbladder without cholecystitis, 43.0%) or choledocholithiasis (calculus of bile duct without cholangitis or cholecystitis without obstruction, 20.5%). Other common diagnoses were calculus of gallbladder with acute cholecystitis (3.5%) calculus of gallbladder with chronic cholecystitis without obstruction (3.2%), calculus of gallbladder and bile duct without cholecystitis without obstruction (7.2%), and other cholelithiasis without obstruction (3.7%). These six diagnoses explained over 80% of the ED revisits for patients with only one biliary-related ED revisit.

F1
Figure 1:
One-year follow-up for symptomatic cholelithiasis patients after index emergency department visit.

There were 39.5% (n = 13,596) of the total cohort that had a cholecystectomy within 1 year and of these, 78.2% (n = 10,627) did not require an ED revisit for a biliary-related diagnosis prior to surgery, while 21.8% (n = 2,969) required one or more ED revisits for a biliary-related diagnosis prior to surgery. Of the total cohort, 10.2% (n = 3,512) had ED revisits within the follow-up period but did not undergo surgery at an OSHPD facility. Half of the population (50.3%, n = 17,319) never returned for a biliary-related diagnosis nor received cholecystectomy within one-year.

Figure 2A shows the distribution of days to cholecystectomy. While over half of patients (62.9%, n = 8,558) who received surgery were operated on within 60 days from their initial ED visit for SC, some received cholecystectomy up to a year out. Figure 2B shows the distribution of days to first ED revisit, with approximately half of patients who returned to the ER coming back within 60 days, but some returning up to 1 year out.

F2
Figure 2:
Time to cholecystectomy and ED revisit for symptomatic cholelithiasis patients after index emergency department visit.

ED Revisits

On bivariate analysis, there were differences in ED revisits after initial discharge based on race/ethnicity and insurance (p < 0.001, Table 1). For example, compared with the 12.6% of non-Hispanic White patients who required ED revisits, 21.2% of Hispanic, and 22.5% of non-Hispanic patients had ED revisits (Table 1). In all three multivariable models, non-Hispanic Black patients had higher odds for recurrent ED use (fully adjusted odds ratio [aOR], 1.23; 95% confidence interval [CI], 1.09–1.39) compared with non-Hispanic White patients. Non-Hispanic Asian and Other patients had significantly reduced odds, and Hispanic patients did not have adjusted differences in requiring recurrent ED use (Table 2) compared with non-Hispanic White patients. In the fully adjusted models, nonprivately insured patients had higher odds for recurrent ED use compared with privately insured patients, with Medicaid patients having the highest odds (aOR, 1.62; 95% CI, 1.47–1.78, Table 2). Age was associated with ED revisits with lower odds of ED revisits as age increases continuously (bivariate analysis and across all three models, Tables 1 and 2). A separate subgroup analysis was performed to specifically examine patients 65 years or older (as a categorical variable) which found that older adults had almost half the odds of repeat ED revisits on fully adjusted models (aOR, 0.59; 95% CI, 0.52–0.67).

On bivariate analysis, there were differences in ED revisits after initial discharge based on race/ethnicity and insurance (p < 0.001, Table 1). For example, compared with the 12.6% of non-Hispanic White patients who required ED revisits, 21.2% of Hispanic, and 22.5% of non-Hispanic patients had ED revisits (Table 1). In all three multivariable models, non-Hispanic Black patients had higher odds for recurrent ED use (fully adjusted odds ratio [aOR], 1.23; 95% confidence interval [CI], 1.09–1.39) compared with non-Hispanic White patients. Non-Hispanic Asian and Other patients had significantly reduced odds, and Hispanic patients did not have adjusted differences in requiring recurrent ED use (Table 2) compared with non-Hispanic White patients. In the fully adjusted models, nonprivately insured patients had higher odds for recurrent ED use compared with privately insured patients, with Medicaid patients having the highest odds (aOR, 1.62; 95% CI, 1.47–1.78, Table 2). Age was associated with ED revisits with lower odds of ED revisits as age increases continuously (bivariate analysis and across all three models, Tables 1 and 2). A separate subgroup analysis was performed to specifically examine patients 65 years or older (as a categorical variable) which found that older adults had almost half the odds of repeat ED revisits on fully adjusted models (aOR, 0.59; 95% CI, 0.52–0.67).

TABLE 2 - Multivariable Regression Model for the Need for Emergency Department Revisits in an OSHPD Population With Symptomatic Gallstones*
Model 1
Patient Characteristics
Model 2
Patient and Geographic Characteristics
Model 3
Patient, Geographic and Hospital Characteristics
OR 95% CI OR 95% CI OR 95% CI
Patient Characteristics
Race/ethnicity (reference: White)
 Hispanic 1.08 0.99–1.17 0.98 0.90–1.07 0.98 0.90–1.06
 Black or African American 1.28 1.131.46 1.23 1.071.40 1.23 1.091.39
 Asian 0.74 0.650.86 0.75 0.640.86 0.76 0.650.88
 Other 0.85 0.71–1.00 0.84 0.701.00 0.83 0.690.99
Insurance (reference: private)
 Medicare 1.36 1.181.56 1.32 1.151.52 1.30 1.131.49
 Medicaid 1.68 1.521.86 1.63 1.491.80 1.62 1.471.78
 Self-Pay 1.39 1.201.61 1.36 1.181.57 1.35 1.171.56
 Other 1.45 1.052.02 1.46 1.072.00 1.47 1.082.01
Age, per 1 y 0.98 0.980.98 0.98 0.980.98 0.98 0.980.98
Sex: female (reference: male) 0.97 0.91–1.04 0.97 0.91–1.03 0.97 0.91–1.03
Geographic characteristics
Zip code income**, per $10,000 0.98 0.960.99 0.98 0.961.00
Bilingual zip code (reference: not bilingual) 1.15 1.071.23 1.15 1.071.24
PCP ratio (reference: above median) 1.16 1.061.27 1.13 1.041.24
Hospital characteristics
Hospital control (reference: investor)
 County/City 0.87 0.70–1.07
 Nonprofit 0.92 0.83–1.01
 District 1.16 0.98–1.36
Hospital size (licensed beds), per 100 beds 1.00 1.00–1.00
Safety-net status (reference: no) 1.13 1.001.26
Teaching hospital (reference: no) 0.95 0.86–1.04
ED volume (total annual visits), per 10,000 visits 1.00 0.99–1.01
*All models performed using logistic regression with standard errors clustered by hospital, bolded values indicate significant association (p < 0.05).
**Median income in the patient's zip code.
Bilingual status defined as >50% of zip code speaking a language other than English.
Ratio of number of PCPs to total population in the patient's county.

Geographic characteristics played a role in explaining variation in recurrent ED utilization. Some odds were attenuated when adding in these variables, for example, the aOR for non-Hispanic Black patients decreased from 1.28 to 1.23 (Table 2). Presentation to a safety-net status hospital was the only hospital characteristic associated with recurrent ED utilization (higher odds: aOR, 1.13; 95% CI, 1.00–1.26, Table 2).

High and Low ED Utilization

Among patients who required an ED revisit for biliary-related problems (n = 6,481) after initial discharge, 27.6% (n = 1,788) were categorized as high utilizers (requiring two or more visits). On univariate analysis, low versus high ED utilization varied by insurance and race/ethnicity status, with non-Hispanic Black (35.9%) and Medicaid (32.3%) patients having the greatest proportion of high ED utilizers as compared with non-Hispanic White (25.5%) or privately insured (21.5%) patients (p < 0.001, Table 3).

TABLE 3 - Patient and Hospital Characteristics for OSHPD Population With Symptomatic Gallstones by Low or High ED Utilization
Variables Total Low ED Utilizer (1 Visit) High ED Utilizer (≥ 2 Visits) p
6,481 (100%) 4,693 (72.4%) 1,788 (27.6%)
Patient characteristics
Race/ethnicity, n (%)* < 0.001
 White 1,861 (28.7%) 1,387 (74.5%) 474 (25.5%)
 Hispanic 3,703 (57.1%) 2,650 (71.6%) 1,053 (28.4%)
 Black or African American 484 (7.5%) 310 (64.1%) 174 (35.9%)
 Asian 218 (3.4%) 183 (83.9%) 35 (16.1%)
 Other 215 (3.3%) 163 (75.8%) 52 (24.2%)
Insurance, n (%)* < 0.001
 Private 2,241 (34.6%) 1,759 (78.5%) 482 (21.5%)
 Medicare 349 (5.4%) 279 (79.9%) 70 (20.1%)
 Medicaid 3,473 (53.6%) 2,353 (67.7%) 1,120 (32.3%)
 Self-pay 364 (5.6%) 262 (72.0%) 102 (28.0%)
 Other 54 (0.8%) 40 (74.1%) 14 (25.9%)
Age, median [IQR] 33 [25, 48] 35 [26, 50] 31 [24, 43] < 0.001
Sex (female), n (%)** 4,927 (76.0%) 3,552 (75.7%) 1,375 (76.9%) 0.31
Geographic characteristics
Zip code income, median [IQR] 59,776 [$47,037, $74,050] $60,732 [$47,553, $75,142] $57,225 [$46,051, $71,586] p < 0.001
Bilingual zip code, n (%)** 2,579 (39.8%) 1,793 (38.2%) 786 (44.0%) < 0.001
PCP ratio§ below median, n (%)** 4,637 (71.6%) 3,255 (69.4%) 1,382 (77.3%) < 0.001
Hospital Characteristics
Hospital control, n (%)* < 0.01
 Investor 1,172 (18.1%) 803 (68.5%) 369 (31.5%)
 County/city 403 (6.2%) 285 (70.7%) 118 (29.3%)
 Nonprofit 4,433 (68.4%) 3,261 (73.6%) 1,172 (26.4%)
 District 473 (7.3%) 344 (72.7%) 129 (27.3%)
Hospital size (licensed beds), median [IQR] 291 [193–420] 291 [195–421] 302 [184–416] 0.14
Safety-net status (yes), n (%) 1,949 (30.1%) 1,318 (28.1%) 631 (35.3%) < 0.001
Teaching hospital (yes), n (%)** 2,675 (41.3%) 1,951 (41.6%) 724 (40.5%) 0.43
ED volume (total annual visits), median [IQR] 69,726 [45,827–89,791] 69,726 [46,395–90,739] 68,748 [45,280–88,095] 0.08
*Percentages for race/ethnicity, insurance, and hospital control represent row percentages (e.g., 74.5% of non-Hispanic White patients were low ED utilizers out of all non-Hispanic White patients with an ED revisit) with denominators shown on the same row in “total” column (percentages in the Total column reflect column percentages).
** Percentages for Sex, PCP Ratio, Safety-Net Status, and Teaching Hospital Status represent column percentages (e.g., 75.6% of all patients who were low ED utilizers were female) with denominators shown at the top of each column.
Median income in the patient's zip code.
Bilingual status defined as >50% of zip code speaking a language other than English.
§Ratio of number of PCPs to total population in the patient's county.

Across all three multivariable models, non-Hispanic Black and non-Hispanic Asian patients had significantly different odds of having two or more ED visits as compared with non-Hispanic White patients (Table 4). Non-Hispanic Black patients had 1.48 the odds of having more frequent ED revisits than non-Hispanic White patients (p < 0.001) on fully adjusted models. Non-Hispanic Asian patients had 0.62 the odds of having more frequent ED revisits than non-Hispanic Asian patients (p < 0.05). Medicaid was the only insurance category that had greater odds for high ED utilization across all three models (fully adjusted: OR, 1.56; 95% CI, 1.34–1.81). In two of three, self-pay patients had higher odds for being high ED utilizers (Model 3: aOR, 1.29; 95% CI, 1.00–1.67). Age was associated with having two or more ED visits with lower odds of higher ED utilization as age increases continuously (bivariate analysis and across all three models, Tables 3 and 4). A separate subgroup analysis was performed to specifically examine patients 65 years or older (as a categorical variable) which found that older adults had almost half the odds of higher ED utilization on fully adjusted models (aOR, 0.61; 95% CI, 0.44–0.85).

TABLE 4 - Multivariable Regression Model for High Versus Low Emergency Department Utilization in an OSHPD Population With Symptomatic Gallstones*
Model 1
Patient Characteristics
Model 2
Patient and Geographic Characteristics
Model 3
Patient, Geographic and Hospital Characteristics
OR 95% CI OR 95% CI OR 95% CI
Patient characteristics
Race/ethnicity (reference: White)
 Hispanic 1.01 0.88–1.16 0.91 0.79–1.05 0.90 0.78–1.04
 Black or African American 1.55 1.251.92 1.50 1.221.85 1.48 1.201.82
 Asian 0.62 0.420.90 0.61 0.410.90 0.62 0.420.91
 Other 0.87 0.64–1.19 0.86 0.63–1.18 0.85 0.62–1.16
Insurance (reference: private)
 Medicare 1.20 0.88–1.63 1.21 0.89–1.66 1.20 0.88–1.64
 Medicaid 1.54 1.321.79 1.56 1.341.82 1.56 1.341.81
 Self-pay 1.28 0.99–1.64 1.29 1.00–1.66 1.29 1.00–1.67
 Other 1.16 0.58–2.32 1.14 0.60–2.18 1.16 0.61–2.19
Age, per 1 y 0.99 0.98–0.99 0.98 0.98–0.99 0.98 0.98–0.99
Geographic characteristics
Zip code income**, per $10,000 1.01 0.98–1.04 1.02 0.99–1.05
Bilingual zip code (reference: not bilingual) 1.24 1.08–1.42 1.24 1.08–1.43
PCP ratio (reference: below median) 1.40 1.18–1.65 1.35 1.14–1.60
Hospital characteristics
Hospital control, (reference: investor)
 County/city 0.79 0.60–1.03
 Nonprofit 0.84 0.71–0.98
 District 0.92 0.70–1.22
Hospital size (licensed beds), per 100 bed 1.00 1.00–1.00
Safety-net status (reference: no) 1.15 1.00–1.33
ED volume (total annual visits), per 10,000 visits 1.00 0.97–1.02
*All models performed using logistic regression with standard errors clustered by hospital, bolded values indicate significant association (p < 0.05).
**Median income in the patient's zip code.
Bilingual status defined as >50% of zip code speaking a language other than English.
Ratio of number of PCPs to total population in the patient's county.

Geographic characteristics such as bilingual make-up of the zip-code (fully adjusted: aOR, 1.24; 95% CI, 1.08–1.43) and county PCP to population ratio (fully adjusted: aOR, 1.35; 95% CI, 1.14–1.60), and hospital control (nonprofit with decreased odds compared with investor-owned: aOR, 0.84; 95% CI, 0.71–0.98) explained some of the variation in high versus low ED utilization (p < 0.05, Table 4).

DISCUSSION

Half of the patients with SC did not have biliary-related revisits within 1 year (following their index ED diagnosis of gallstones). Non-Hispanic Black and not privately insured patients were more likely to revisit the ED for biliary complaints and this was consistent for low versus high ED utilizers. These results can help guide the current understanding of SC and inform future research regarding standardization of care for this common surgical problem.

Based on our findings that a large proportion of patients with SC (who did not undergo initial cholecystectomy) do not return to the ED or require surgery, patient counseling should center around anticipatory guidance regarding future attacks. However, the role and optimal timing of surgical intervention for SC have yet to be determined. Previous trials that have compared surgery to observation for patients with SC examined patients at varied stages of disease presentation. One trial including patients who had a range of zero to over five biliary colic attacks a month.19,20 Our results suggest that future studies should focus on advanced disease presentation (e.g., randomizing patients to surgery versus observation after their second or third biliary colic attack) to understand when the risk of surgery outweighs the negative consequences of repeat ED visits or progression of disease. While our study shows that many patients did not require further resource utilization, our analyses took a closer look at the cohort of patients who did require return ED visits after their initial presentation with SC.

The patients with SC who had higher risk for the occurrence of an ED revisit and more frequent ED visits for biliary complaints were non-Hispanic Black, uninsured or Medicaid. Living in bilingual zip-codes or counties with a lower PCP to population ratio was also associated with occurrence of and more frequent ED revisits. Our findings suggest that already vulnerable patients are disproportionately experiencing the burden of increased health care resource utilization. This burden can result in a perpetuating cycle, in which frequent ED revisits lead to disadvantaged patients experiencing additional financial impacts such as lost time at work or disruption of childcare.21 Notably, while other measures of access to care (e.g., insurance, acculturation) explained variations in ED utilization for Hispanic patients, differential ED utilization for non-Hispanic Black patients persisted in fully adjusted models. There is a large body of literature showing that non-Hispanic Black patients in particular, experience surgical disparities in this country.22–25 Our results add to these data suggesting that implicit biases in medicine integrate deeply within the structure of health care. While research has largely focused on disparities in outcomes such as mortality, length of stay or postoperative complications, this work implies that these problems may originate further upstream.22,26 Understanding and improving the efficiency of how patients use health care for SC may ensure equitable treatment, and improve outcomes.

One way to potentially decrease differences in patient utilization of ED resources is to implement standardized protocols regarding follow-up, specifically targeting high utilizers. For example, at one tertiary care center, a scheduling assistance program increased rate of PCP follow-up and decreased readmissions after inpatient hospital admissions.27 At integrated health systems, generating an automated PCP appointment after an ED visit for SC to ensure comprehensive discharge planning could improve timely access to follow-up, with the goal of reducing unnecessary ED visits. The problem becomes more complicated for uninsured patients. Our study found that nonprivately insured patients were at risk for greater ED utilization, highlighting the consequences of disparate access to care, particularly on already marginalized populations. Prior work has shown that expanding insurance coverage can impact a surgical population, with improved access to multiple surgical treatments after Medicaid expansion, compared with nonexpansion states.28–30 Specifically, there was a 9.8% increase in cholecystectomies in expansion versus nonexpansion states at the facility level. Until equitable access to health care is achieved, automatically generated PCP appointments and standardized follow-up protocols will only benefit a select population of patients. In the meantime, efforts can focus on education during ED discharge. Patients should receive a set of discharge instructions in their own language about what to expect with a diagnosis of symptomatic cholelithiasis, when to return to the ED, and when to watchfully manage their symptoms. Guidelines should clearly indicate when surgery is recommended for patients with SC (e.g., after two or more return ED visits for pain attacks) to prevent patients from falling into the high ED utilization category. Such standardization of education and guidelines will help facilitate equitable care, regardless of race/ethnicity.

This study has several limitations. First, OSHPD is an administrative database, and lacks some clinical nuances. The use of ZCTAs to capture geographic variables was limited by abrupt cutoffs between regions, losing the granularity of subtle changes between regions. These boundaries do not capture dynamic trends in gentrification and only allow for summary variables within a region, presenting a more static analysis. In addition, topological features may lead to blunting of geospatial analysis with the use of ZCTAs. The cohort was identified using restrictive ICD codes to isolate a population with only SC and without more advanced disease pathology at index ED presentation. However, it is possible that patients incorrectly coded with SC were included or patients with SC that did not meet the ICD code criteria were missed. Second, we did not capture data from PCPs or ambulatory surgery centers not in OSHPD. Patient's interactions with their PCP before or after ED visits for SC is a central consideration for future work. Finally, the EDD does not contain comorbidity data, so we were unable to adjust for this, since many patients were only recorded in the EDD dataset. Despite these limitations, we felt that these results are generalizable due to the diverse population captured by this dataset of over nearly 40 million individuals with hospitals ranging in size and location.

This comprehensive analysis found that non-Hispanic Black patients experience higher utilization of health care resources for SC, suggesting that we must find strategies to mitigate these disparities. One such idea is the development of standardized protocols regarding the follow-up and education for SC. To generate these clinical recommendations, dedicated trials should examine the best timing of surgery for SC across the spectrum of disease presentation. Furthermore, work will also be needed to assess barriers to receiving care for these patients in order to broadly improve equitable access to treatment. In doing so, we can work toward reducing disparities in resource utilization and clinical outcomes for this common surgical disorder.

AUTHORSHIP

All authors substantially contributed to the conception and design of the study. R.S., M.D., and P.K. contributed to the literature search and data collection. R.S. and N.J. contributed to data analysis. R.S., N.J., M.M.-G., M.M.R., and D.Z. contributed to the data interpretation. R.S. and M.G. drafted the article, and all authors critically revised it for important intellectual content. All authors gave final approval for this version to be published.

ACKNOWLEDGMENT

R.S. was supported by the VA Office of Academic Affiliations through the VA/National Clinician Scholars Program during their time on this project.

DISCLOSURE

The authors declare no conflicts of interest.

REFERENCES

1. Shaheen NJ, Hansen RA, Morgan DR, Shaheen NJ, Hansen RA, Morgan DR, et al. The burden of gastrointestinal and liver diseases, 2006. Am J Gastroenterol. 2006;101(9):2128–2138.
2. Sandler RS, Everhart JE, Donowitz M, Adams E, Cronin K, Goodman C, et al. The burden of selected digestive diseases in the United States. Gastroenterology. 2002;122(5):1500–1511.
3. Gurusamy KS, Koti R, Fusai G, Davidson BR. Early versus delayed laparoscopic cholecystectomy for uncomplicated biliary colic. Cochrane Database Syst Rev. 2013;(6):Cd007196.
4. Attili AF, de Santis A, Capri R, Repice AM, Maselli S. The natural history of gallstones: the GREPCO experience. Hepatology. 1995;21(3):656–660.
5. McSherry CK, Ferstenberg H, Calhoun WF, Lahman E, Virshup M. The natural history of diagnosed gallstone disease in symptomatic and asymptomatic patients. Ann Surg. 1985;202(1):59–63.
6. Gurusamy KS, Davidson C, Gluud C, Davidson BR. Early versus delayed laparoscopic cholecystectomy for people with acute cholecystitis. Cochrane Database Syst Rev. 2013;(6):CD005440.
7. Glasgow RE, Cho M, Hutter MM, Mulvihill SJ. The spectrum and cost of complicated gallstone disease in California. Arch Surg. 2000;135(9):1021–1025.
8. Dimou FM, Adhikari D, Mehta HB, Riall TS. Trends in follow-up of patients presenting to the emergency department with symptomatic cholelithiasis. J Am Coll Surg. 2016;222(4):377–384.
9. Altieri MS, Yang J, Zhu C, Sbayi S, Spaniolas K, Talamini M, et al. What happens to biliary colic patients in New York state? 10-year follow-up from emergency department visits. Surg Endosc. 2018;32(4):2058–2066.
10. Tang N, Stein J, Hsia RY, Maselli JH, Gonzales R. Trends and characteristics of US emergency department visits, 1997-2007. JAMA. 2010;304(6):664–670.
11. Mayfield CA, Geraci M, Dulin M, Eberth JM, Merchant AT. Social and demographic characteristics of frequent or high-charge emergency department users: a quantile regression application. J Eval Clin Pract. 2021;27(6):1271–1280.
12. Giannouchos TV, Kum HC, Foster MJ, Ohsfeldt RL. Characteristics and predictors of adult frequent emergency department users in the United States: a systematic literature review. J Eval Clin Pract. 2019;25(3):420–433.
13. Data from: Emergency Department, Patient Discharge and Ambulatory Surgery Databases. 2016–2018. California. https://oshpd.ca.gov/data-and-reports/.
14. Data from: United States Census Bureau Data.
15. Data from: American Medical Association Physician Masterfile 2016.
16. Data from: OSHPD Hospital Annual Financial Disclosure Reports 2016.
17. Reiter KL, Jiang HJ, Wang J. Facing the recession: how did safety-net hospitals fare financially compared with their peers?Health Serv Res. 2014;49(6):1747–1766.
18. Sutton JP, Washington RE, Fingar KR, Elixhauser A. Characteristics of safety-net hospitals, 2014: statistical brief# 213. 2016.
19. Vetrhus M, Soreide O, Solhaug JH, Nesvik I, Sondenaa K. Symptomatic, non-complicated gallbladder stone disease. Operation or observation? A randomized clinical study. Scand J Gastroenterol. 2002;37(7):834–839.
20. Schmidt M, Søndenaa K, Vetrhus M, Berhane T, Eide GE. A randomized controlled study of uncomplicated gallstone disease with a 14-year follow-up showed that operation was the preferred treatment. Dig Surg. 2011;28(4):270–276.
21. Haider AH, Obirieze A, Velopulos CG, Richard P, Latif A, Scott VK, et al. Incremental cost of emergency versus elective surgery. Ann Surg. 2015;262(2):260–266.
22. Haider AH, Scott VK, Rehman KA, Velopulos C, Bentley JM, Cornwell EE III, et al. Racial disparities in surgical care and outcomes in the United States: a comprehensive review of patient, provider, and systemic factors. J Am Coll Surg. 2013;216(3):482–492.e12.
23. Lucas FL, Stukel TA, Morris AM, Siewers AE, Birkmeyer JD. Race and surgical mortality in the United States. Ann Surg. 2006;243(2):281–286.
24. Ricciardi R, Selker HP, Baxter NN, Marcello PW, Roberts PL, Virnig BA. Disparate use of minimally invasive surgery in benign surgical conditions. Surg Endosc. 2008;22(9):1977–1986.
25. Lee SL, Yaghoubian A, Stark R, Shekherdimian S. Equal access to healthcare does not eliminate disparities in the management of adults with appendicitis. J Surg Res. 2011;170(2):209–213.
26. Torain MJ, Maragh-Bass AC, Dankwa-Mullen I, Hisam B, Kodadek LM, Lilley EJ, et al. Surgical disparities: a comprehensive review and new conceptual framework. J Am Coll Surg. 2016;223(2):408–418.
27. Marcondes FO, Punjabi P, Doctoroff L, Tess A, O'Neill S, Layton T, et al. Does scheduling a postdischarge visit with a primary care physician increase rates of follow-up and decrease readmissions?J Hosp Med. 2019;14:E37–e42.
28. Al-Refaie WB, Zheng C, Jindal M, Clements ML, Toye P, Johnson LB, et al. Did pre-affordable care act Medicaid expansion increase access to surgical cancer care?J Am Coll Surg. 2017;224(4):662–669.
29. Giladi AM, Aliu O, Chung KC. The effect of Medicaid expansion in New York state on use of subspecialty surgical procedures by Medicaid beneficiaries and the uninsured. J Am Coll Surg. 2014;218(5):889–897.
30. Lin S, Brasel KJ, Chakraborty O, Glied SA. Association between Medicaid expansion and the use of outpatient general surgical care among US adults in multiple states. JAMA Surg. 2020;155(11):1058–1066.
Keywords:

Symptomatic cholelithiasis; cholecystectomy; utilization; emergency department; disparities

Supplemental Digital Content

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.