Introduction

Diverticular disease of the colon is one of the most prevalent conditions in western society and is a leading cause of both outpatient visits and hospitalizations.1 Diverticulitis, the acute inflammation and/or infection of the colonic diverticulum, is commonly an emergent condition.2 Diverticulitis affects 70 % of Americans by age 80, and healthcare-related costs may exceed 4.5 billion dollars annually in the USA.3 Acute diverticulitis is also responsible for around 314,000 annual discharges, 1.5 million days of care, and accounts for an average length of stay of 4.8 days.4

The two mainstays of treatment are broad-spectrum antibiotics along with bowel rest and surgical intervention5 Antibiotic therapy is reserved for initial presentation of uncomplicated acute diverticulitis.6–8 Emergent surgical management is reserved for patients with life-threatening complications,9 while elective surgical management is used in cases of recurrence, and/or when conservative management fails to resolve symptoms10 These management decisions frequently incorporate patient preferences as well.

The management of diverticulitis is often dependent upon physician preferences, patient preference, patient’s medical history, and available resources. One study showed significant national variability among physicians choosing to employ surgical management.11 Other studies have also suggested that race and insurance status were associated with different management patterns for diverticulitis patients, as well as differences in mortality.12 Despite all this data, no correlative trends have been characterized in the state of New York between vulnerable populations (including racial, ethnic, age, gender, and socioeconomic cohorts) and diverticulitis management.

Our study sought to explore and identify any correlative relationships in management of diverticulitis and, if present, analyze their significance, with the potential to re-focus and refine resources, funding, and/or research efforts associated with the care of diverticulitis in New York and the USA.

Materials and Methods

Following approval by our institutional review board (IRB) and the New York Department of Health (DOH), 265,724 patients with the primary diagnosis of diverticulitis from 1995 to 2014 were identified from the New York Statewide Planning and Research Cooperative System (SPARCS) administrative database. SPARCS is a longitudinal comprehensive data reporting system which collects patient level data on patient characteristics, diagnoses and treatments, services, and charges for hospital discharge, ambulatory surgery patient, and emergency department admission in New York State. Data is collected from all Article 28 Facilities in New York State as well hospital-based and free-standing ambulatory surgery facilities. Patients were identified via the SPARCS database using International Classification of Diseases, Ninth Revision (ICD-9) primary diagnosis code for diverticulitis (ICD-9 562.11, 562.13) from both inpatient and outpatient records. Among records with primary diagnosis of diverticulitis, to define if any procedure was performed during that visit, primary procedure column was further searched for ICD9 codes from both inpatient files (1995–2014) and outpatient files (1995–2007) (Table 1), and seven CPT columns were searched for CPT codes from outpatient files (2008–2014) (Table 2). Among these records, there were 2788 in-hospital deaths. Patients younger than 18 or without complete records were excluded from the study population.

Table 1 Defined procedures and ICD-9 procedure codes for inpatient (1995–2014) and outpatient (1995–2007)
Full size table
Table 2 Defined procedures and procedure codes for outpatient (2008–2014)
Full size table

Primary procedures were defined as either emergent or elective. If the type of inpatient admission was classified as emergent or urgent in SPARCS database and that patient went on to have a procedure performed during that admission, the surgery type was defined as emergent for the purposes of this study. If the type of inpatient admission was classified as anything else and the patient went on to have a procedure performed during that admission, the surgery type was defined as elective. For all outpatient procedures performed, the surgery type was defined as elective.

Chi-square test and Welch’s t test were used to compare categorical variable and continuous variables, respectively. Logistic regression was performed to compare linear trends of emergent surgery percentage within each subgroups of age group, gender, payment type, and race/ethnicity. Trends for the percentage of each subgroup were illustrated by the estimated odds ratio along with their 95 % confidence intervals. All analysis was performed using SAS 9.3 (SAS Institute Inc., Cary, NC), and significance level was set at 0.05. All statistical analyses were performed and reviewed by a statistician (J.Y.).

Results

Table 3 and Fig. 1 show the number of newly diagnosed patients in each year from 1995 to 2014, which increased by 206 % from 8505 patients in 1995 to 17,578 patients in 2014. Table 3 also shows the number of patients who went on to have surgery that year, as well as whether or not their surgery was emergent or elective.

Table 3 Frequency table for diagnosis and surgery records of diverticular disease (1995–2014)
Full size table
Fig. 1
figure 1

Total number of patients who received primary diagnosis of acute diverticulitis from 1995 to 2014

Full size image

Figure 2 shows the trend of the total surgery, emergent, and elective surgery percentages from Table 3. Overall, there is a decreasing trend in the percentage of emergent surgery over time: the odds of having emergent surgery decreased by 3.79 % yearly (OR: 0.9621, 95 % CI: 0.9592–0.9651, p value < 0.0001). In addition, the percentage of newly diagnosed patients who went on to have surgery decreased from around 31 % in 1995 to 10 % in 2014.

Fig. 2
figure 2

Trend for percentage of patients who ever had surgery later on, emergent, and elective surgery

Full size image

Figure 3 shows the percentage of emergent surgery over years by patients’ characteristics such as age (Fig. 3a), gender (Fig. 3b), race/ethnicity (Fig. 3c), and payer (Fig. 3d). Table 4 shows the odds ratio of yearly trends of emergent surgery percentage given by patient’s demographics. The linear trends were significantly different across gender, race/ethnicity, and most payment methods (p values < 0.001). For patients older than 80 years old, the yearly trend is not significant (p value = 0.0738). However, the linear trends were significantly different across all other age groups (p value < 0.001).

Fig. 3
figure 3

Percentage of emergent surgery over years by patients’ characteristics such as age (a), gender (b), race/ethnicity (c), and payer (d)

Full size image
Table 4 Odds ratio of yearly trend of emergent surgery percentage given by patient’s demographics
Full size table

Discussion

This studied showed an incremental increase in nonoperative management over the past 20 years. Our data showed the percentage of patients receiving surgery in 1995 was 31 % as compared to 10 % in 2014 in the setting of increasing cases of diverticulitis. Currently, guidelines suggest bowel rest or intake of oral fluids, and a 7–10-day regimen of broad-spectrum antibiotics is recommended in patients with uncomplicated acute diverticulitis (AD).13 This treatment strategy has been reported to be successful in 85–100 % of patients.14,15 In addition, one study corroborates that nonoperative management is safe in patients with AD because fewer than 5 % of patients will need an emergent procedure should they have a subsequent attack.16 This information is corroborated by the trends we observed. Interestingly, more recently, the AVOD and Cochrane trials have found that antibiotic therapy did not prevent complications, accelerate recovery, or prevent recurrences17, and there were no significant differences in treatment or complications between antibiotics and no antibiotics for the treatment of uncomplicated diverticulitis.18 These findings suggest the use of an even more conservative approach consisting of intravenous hydration and bowel rest. With such debate, it was important to explore if clinical management trends mirrored the findings of recent studies.

In addition, our study found that management is trending towards elective surgery, with a decreasing percentage of emergent surgeries being performed. These shifts could reflect the superior antibiotic therapy as well as recent management parameters suggesting elective procedures following two AD attacks. When to employ operative management on a patient suffering from AD has been a point of controversy for many years. Some studies concluded that the risk of emergency surgery increases with increasing number of recurrent attacks,19 while others found that the risk of free perforation decreased with increasing number of previous episodes.20 The decision to perform elective surgery to prevent recurrence has been controversial as well. Some studies suggest a patient’s medical condition (i.e., age, comorbidities, symptoms, etc.) should influence the decision to perform elective procedures more so than the number of AD attacks,21 while others find that performing elective surgery after two attacks significantly decreases morbidity and mortality.22 Another line of argument is that elective surgery should be performed on patients under the age of 50 whom experience AD due to their longer life span and, thus, increased risk of reoccurrence,23 but subsequent studies have found no significant differences in reoccurrence rates within this younger patient subgroup.24,25

Finally, our study explored whether or not these management trends were reflected across gender, race/ethnicity, payment type, and age subgroups. In all the subgroups besides payment subgroups “other” and “unknown” and age subgroup “>80,” there was a significant decrease in the odds a patient received emergent surgery. This argues that management paradigms have changed equally regardless of patient demographics and socioeconomic status which is important when monitoring healthcare equality. Within the older “>80” age group, the trend was still towards nonoperative management, albeit not statistically significant, which could reflect the decreased likelihood of performing surgery on patients in that age group.

The strengths of the study include the sample size, dataset used, and years of analysis. SPARCS is an ideal database to capture admissions as it tracks data across all participating New York hospitals. By capturing all procedures and grouping them into emergent or elective categories, trends of nonoperative management in the past 20 years were able to be fully classified in the state of New York. Limitations of this study include all those that accompany use of an administrative database, namely coding errors as well as inability to access clinically rich data. Severity of disease, imaging, antibiotic choice, and duration of treatment were unable to be accounted for.

Conclusion

Over the past 20 years in New York State, there has been an increase in the prevalence of diverticulitis diagnoses and hospital admissions. Interestingly, there has been a significant shift toward nonoperative management, in particular at initial presentation as is represented by the decrease in percentage of patients receiving operative management. In addition, the percentage of patients within the operative management arm receiving emergent surgeries has decreased while elective surgeries has increased during this time. These trends were displayed across most gender, race/ethnic, payment type, and age subgroups with the notable exception of patients over 80 years old. The shift in management likely reflect more effective antibiotic treatment as well as changes in management paradigms favoring a more conservative approach. In order to more fully assess the benefits of operative vs. nonoperative management, longitudinal outcome analysis of the past 20 years can help shed light on whether or not this shift has resulted in superior patient care, as well provide information to help define more concrete management protocols to ensure safe and effective treatment of acute diverticulitis.