Comorbidity Patterns and Healthcare Resource Utilization in COPD Patients: A Retrospective Analysis Using Association Rules

doi:10.21203/rs.3.rs-4901646/v1

Download PDF

Research Article

Comorbidity Patterns and Healthcare Resource Utilization in COPD Patients: A Retrospective Analysis Using Association Rules

https://doi.org/10.21203/rs.3.rs-4901646/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Patients with chronic obstructive pulmonary disease (COPD) frequently present with multiple concurrent chronic diseases, which has a profound impact on their quality of life. At present, there is a paucity of research investigating the interrelationships between COPD and comorbidities. Furthermore, there is a dearth of systematic and comprehensive studies in this area.

Methods

The data were extracted from the discharge summaries of patients whose primary diagnosis was COPD (ICD-10: J44). The 15 most frequent comorbidities in patients’ other diagnoses were identified, and the association rule mining (ARM) method with the apriori algorithm was then employed to derive comorbidity combinations associated with length of stay (LOS) and antimicrobial drug costs.

Results

The study encompassed data from 2,159 patients. The five most prevalent comorbidities were other respiratory diseases (41.69%), hypertension (31.91%), liver diseases (25.34%), heart diseases (20.84%), and arrhythmia (15.05%). Some comorbidity combinations identified by ARM were found to be significantly associated with longer LOS of at least seven days. Conversely, some were associated with higher antimicrobial drug costs. The combination of diabetes, heart disease, other respiratory disease, and hypertension demonstrated the strongest association with longer LOS (adjusted OR (aOR): The odds ratio was 3.03 (95% CI: 1.03–8.88). The combination of diabetes, other respiratory diseases and hypertension was associated with higher antimicrobial drug costs, amounting to a minimum of 400 RMB (aOR: 2.63, 95% CI: 1.31–5.31).

Conclusion

Other respiratory diseases and hypertension were among the most frequently reported comorbidities in patients with COPD in Guangzhou, China. It is possible that specific groups of comorbidities may contribute to a greater burden of healthcare resource utilization. Further investigation is required to elucidate the mechanisms underlying these associations and to identify potential measures to optimize the healthcare resource utilization.

COPD

comorbidity patterns

healthcare resource utilization

association rules

apriori algorithm

Chronic obstructive pulmonary disease (COPD) is defined as a heterogeneous lung disease characterized by chronic respiratory symptoms (dyspnea, cough, sputum and worsening of breathlessness) due to airway abnormalities (bronchiectasis, bronchiectasis minor) and alveolar abnormalities (emphysema) resulting in persistent, progressive airflow obstruction[1]. COPD is a highly prevalent chronic disease, currently the third leading cause of death worldwide[2] and a major public health challenge[3]. The prevalence of COPD in China is as high as 8.2% in people over 40 years old, which has a significant impact on the health and quality of life of those affected[4]. Comorbidity is used to describe the additional burden of disease that coexists with a specific disease[5]. Although COPD is primarily a pulmonary disease, it is now also recognized as a complex multicomponent disease characterized by chronic systemic inflammation, often in combination with other chronic diseases[6]. A review of the literature indicates that 84.5% of patients with COPD have at least one comorbidity[7]. The interaction between COPD and comorbidities may result in an accelerated disease progression, a reduction in the quality of life of patients and the survival rates of patients[8–10].

Although many studies have investigated the impact of comorbidities or multi-morbidities on healthcare resource utilization[11], only a few studies focused on the comorbidities in COPD patients. A systematic review found that COPD patients with comorbidities had higher health care expenditures[12]. This high demand for healthcare services not only affects the quality of life of patients, but also puts pressure on the healthcare system. Therefore, optimizing management strategies for COPD and its comorbidities and adopting a more systematic and multidisciplinary approach are important to improve patient outcomes and reduce healthcare resource consumption.

ARM is a data-driven approach to identify frequent item sets in large transaction databases[13]. Recently, it can be used for comorbidity patterns[14, 15]. Compared to previous comorbidity analysis methods, ARM can efficiently handle millions of possible combinations of comorbidities and identify combinations that occur frequently or correlate with a given outcome[16]. Although many studies using ARM methods have examined patterns of comorbidity in patients with a variety of primary diseases[14], in the general population[17], or in elderly patients[18], few studies have focused on COPD comorbidities or examined the link between patterns of comorbidity and healthcare resource utilization.

The aim of this study was to explore the pattern of comorbidities in COPD patients using ARM and to investigate the association between different combinations of comorbidities and LOS and antimicrobial drug costs.

2.1 Selected objects

This was a single-center retrospective study and a total of 2159 patients were included. COPD patients admitted to the First Hospital of Guangzhou Medical University from 1 January 2023 to 31 December 2023 were selected for the study. Inclusion criteria: (1) primary discharge diagnosis consistent with COPD diagnosis (2) age greater than or equal to 40 years; Exclusion criteria: (1) patients with lung transplantation. This study has been approved by the Ethics Committee, approval number: ES-2024-K085-01.

2.2 Methods

2.2.1. Retrospective study: The discharge data included patients’ socio-demographic information, inspection indicators, hospitalization expenses as well as the primary diagnosis and secondary diagnoses (coded according to the International Classification of Disease 10th edition (ICD-10). Among them, age was divided into 5 groups: 40–49, 50–59, 60–69, 70–79, and ≥ 80 years; educational level was divided into 3 groups: primary school and below, junior and senior high school, and college and above; BMI was divided into 4 groups: <18. 5kg/m², 18. 5-25kg/m², 25-30kg/m², and ≥ 30kg/m²; and Autonomy for Daily Living (ADL) was divided into 4 groups according to the scores, of which 100 was classified as grade A, indicating complete self-care, 60–99 was classified as grade B, indicating mild dependence, 40–59 was classified as grade C, indicating moderate dependence, and < 40 was classified as grade D, indicating severe dependence[19]. In the case of multiple admissions during the year, the last admission is counted, and the test indicators are based on the results of the first blood sample taken on admission.

2.2.2 Comorbidity: The comorbidity analyses included in this study covered the following 15 common chronic diseases, all of which met ICD-10 diagnostic criteria Hypertension (I10-I15), Diabetes (E10-E14), Cancer (C00-C96), Heart diseases (including myocardial infarction, coronary heart diseases, angina pectoris and congestive heart failure, I20-I25, I50 and Z95), Arrhythmia (I45-I49, R00), Cerebrovascular diseases (I60-I69), Emotional disorders (F20-F48), Arthropathy (M00-M25), Dyslipidemia (with elevated LDL, triacylglycerol and total cholesterol or low HDL levels, E78), Liver diseases (K70-K77) and Kidney diseases (N18-N23), Digestive diseases (K20-K31), Other respiratory diseases (including asthma, bronchiectasis, pulmonary hypertension, pulmonary heart disease, respiratory failure, J44.807, J45-J47, J96, I27), Memory diseases (dementia, cerebral atrophy, Parkinson's disease, G20, G30, G31, F01-F03) and Sleep disorders (G47).

2.3. Statistical analyses

Data were pre-processed using the R programming language, and missing values were handled using multiple imputation methods to ensure the robustness of the analysis. Specifically, the "mice" package in R was used for multiple imputation to handle missing data. Data were analyzed using SPSS version 26. Categorical data were expressed as number of cases (percentage) [N (%)] and analyzed using the chi-square test. Continuous variables were tested for normality using the Shapiro-Wilk test. For non-normally distributed data, the Mann-Whitney U test was used. Results were reported as median with interquartile range (LQ-UQ). A p-value of less than 0.05 was considered statistically significant.

We used the ARM method with the Apriori algorithm to (1) identify the most common 2-way, 3-way, and 4-way combinations; and (2) determine which combinations were significantly associated with longer LOS and higher antimicrobial drug costs. The Apriori algorithm[20] is designed to find the frequent subsets of items (in this case comorbidities) among a set of transactions (in this case other diagnoses in the patient's index admission records). Describe the mined association rules using three metrics: Support, confidence, and lift. Support: (A→B) = P(AB) in sample database D is the frequency of A and B occurring simultaneously; Confidence: (A→B) = P(B|A) in sample database D is the conditional probability of B occurring conditional on A occurring; Lift: (A→B) = P(B|A)/P(AB) in sample database D is the conditional probability of B occurring under conditions in which A occurs, the conditional probability of B occurring is a multiple of the unconditional probability of B occurring[21]. A high rule support indicates a higher probability of A and B occurring simultaneously in this database. A higher confidence level indicates a higher probability of B occurring when A occurs. And a lift of 1 as the dividing line, when the lift is > 1, it indicates that there is a positive direction of correlation from A to B, which means that the probability of B occurring under the condition of A's existence is increased by a factor of one. These three criteria are indispensable, and the use of any one of them alone may lead to erroneous conclusions[22]. we set the minimum support of 1% and the minimum confidence of 50% to filter out the most important rules. Association rules were applied to logistic regression models for univariate and multivariate logistic regression analyses, respectively. For LOS, we dichotomized as greater than or equal to 7 days, and for antimicrobial drug costs, we dichotomized as greater than or equal to the median.

3.1. Comorbidities in COPD patients

A total of 2159 COPD patients were enrolled in this study and 15 comorbidities were recorded (Fig. 1). The prevalence of comorbidities was listed in Fig. 2. The top five most common comorbidities were other respiratory diseases (46.09%), hypertension (32.19%), liver diseases (25.34%), heart diseases (20.84%) and arrhythmia (15.05%). The mean number of comorbidities (1.930 ± 1.392) was found in 2159 COPD patients. Of these, 1871 had ≥ one comorbidities, representing 86.66% of all patients. The largest number of patients with one comorbidity was 659 (30.52%), 570 patients with two comorbidities (26.40%), 347 patients with three chronic diseases (16.07%) and 295 patients with ≥ four chronic diseases (13.66%), as shown in supplementary file 1.

3.2. Comorbidity in COPD Patients: Demographics, Clinical Characteristics, and Healthcare Resource Utilization

Demographic characteristics showed a significant increase in the prevalence of comorbidities with age, from 66.67% in the 40–49 years group to 90.42% in the 80 years and over group (X²=26.847, p < 0.001). In addition, ADL likewise showed a significant positive correlation with the prevalence of comorbidities. The data showed that the prevalence of comorbidities was 82.11% in patients who were fully self-sufficient, while the rate was as high as 94.90% in patients with severe dependence (X²=22.618, p < 0.001). Regarding BMI, the prevalence of comorbidities in the higher BMI group (≥ 30 kg/m2) reached 97.44%, which was much higher than in the other groups (X²=11.025, p = 0.012), as shown in Table 1. However, factors such as sex, smoking history and education did not show statistically significant differences in the analysis of comorbidities (sex X²=0.196, p = 0.661; smoking history X²=1.801, p = 0.180; education X²=1.357, p = 0.507), as shown in supplementary file 2. The distribution of comorbidities stratified by patient demographic characteristics is shown in a heat map (Fig. 3), where higher prevalence is indicated by darker purple color. We observed that the prevalence of comorbidities such as hypertension, heart diseases, Arrhythmia, cerebrovascular disease, cancer, arthropathy, and memory disorders increased with age. patients with a BMI ≥ 30 kg/m² had a higher prevalence of multiple comorbidities, with hypertension and sleep disorders being more prominent. some comorbidities were more prevalent in those with a lower ADL, with a relatively higher prevalence of other respiratory disorders, heart diseases, cerebrovascular diseases were relatively higher.

Table 1

Impact of Age, BMI, and ADL Classification on Comorbidity Distribution
Characteristics	cases	Number of Comorbidities					Comorbidity rate(%)	X² value	P value
Characteristics	cases	0	1	2	3	≥ 4	Comorbidity rate(%)	X² value	P value
Age (years)								26.847	< 0.001
40–49	27	9	10	6	2	0	18(66.67)
50–59	963	52	92	62	31	26	211(80.23)
60–69	830	118	263	224	144	81	712(85.78)
70–79	778	84	238	220	115	121	694(89.20)
≥ 80	261	25	56	58	55	67	236(90.42)
BMI(kg/m2)								11.025	0.012
< 18.5	583	78	217	156	74	58	505(86.62)
18.5–25	1278	187	375	328	210	178	1091(85.37)
25–30	259	22	57	81	54	45	237(91.50)
≥ 30	39	1	10	5	9	14	38(97.44)
ADL								22.618	< 0.001
A	710	127	219	194	99	71	583(82.11)
B	1117	132	357	288	188	152	985(88.18)
C	234	24	66	67	39	38	210(89.74)
D	98	5	17	21	21	34	93(94.90)

Clinical characteristics showed notable differences between COPD patients with and without comorbidities, in which neutrophil ratio (NR), partial pressure of carbon dioxide (blood gas PCO2), fibrinogen (Fbg), D-dimer (DD), blood glucose (at random) (BG), pro-brain natriuretic peptide (pro-BNP), Procalcitonin (PCT) and triglyceride (TG) were higher in the patients with comorbidities than in the group without comorbidities, and the levels of hemoglobin (Hb), platelet count (PLT), blood gas pH (pH) and albumin (Alb) were lower in the patients with comorbidities than in the group without comorbidities. The values in the group with comorbidities were lower, with p-value less than 0.05 (Table 2).

Table 2

Comparison of clinical characteristics between patients with and without comorbidities
Variable	without comorbidities (n = 288)	with comorbidities (n = 1871)	z-value	p-value
WBC	7.315(5.982, 9.260)	7.500(5.900, 9.750)	-0.748	0.455
NR	67.550 (59.425, 77.200)	71.200 (62.700, 81.300)	-3.767	< 0.001
RBC	132.000 (123.000, 143.000)	131.000 (118.000, 141.000)	-1.982	0.048
PLT	233.000 (194.00,295.00)	223.000 (179.000,280.000)	-2.568	0.010
pH	7.390 (7.362, 7.413)	7.384(7.351, 7.410)	-2.682	0.007
PO2	93.250(78.925, 118.100)	92.400(78.000, 121.600)	-0.060	0.952
PCO2	42.850 (40.100, 47.400)	45.500(40.100, 53.900)	-4.783	< 0.001
FIB	3.655 (3.078, 4.678)	3.910 (3.180, 4.940)	-2.498	0.012
DD2	428.500 (288.250, 852.500)	526.000(326.000, 968.000)	-3.097	0.002
ALT	15.850(12.050, 22.100)	16.700(11.900, 24.500)	-1.440	0.150
AST	20.600(18.200, 25.000)	21.500(17.500, 26.900)	-0.844	0.399
BG	5.675 (4.740, 7.095)	6.220 (5.070, 8.170)	-4.986	< 0.001
SCR	77.250(65.725, 87.525)	77.000(65.900, 91.800)	-0.623	0.533
pro-BNP	69.000 (35.900, 143.250)	112.000 (50.000, 298.000)	-6.181	< 0.001
PCT	0.000 (0.000,0.000)	0.000(0.000,0.050)	-2.051	0.040
ALB	36.950 (34.225, 39.900)	36.300 (33.600, 38.900)	-2.559	0.010
TG	0.850 (0.643, 1.160)	0.860(0.650,1.200)	-0.968	0.333
TC	4.605(3.990, 5.668)	4.660(3.910, 5.550)	-0.289	0.773
HDL	1.320(1.100, 1.570)	1.290(1.090, 1.550)	-1.029	0.304
LDL	2.920(2.443, 3.520)	2.960(2.400, 3.610)	-0.088	0.930
ESR	22.000(9.000, 45.000)	24.000(9.000, 51.000)	-1.141	0.254

Healthcare resource utilization showed differences in medical costs and hospitalization data between COPD patients with and without comorbidities. The total cost of hospitalization, the cost of western medicines, the cost of antimicrobials, the number of days of hospitalization, the number of hospitalizations for COPD in the past year and the number of previous hospitalizations were higher in patients with comorbidities compared with those without comorbidities. The differences were statistically significant with p values less than 0.05, as shown in Table 3.

Table 3

Comparison of medical costs and hospitalization data between patients with and without comorbidities.
Variable	Without Comorbidities (n = 288)	With Comorbidities (n = 1871)	Z-value	P-value
Total Cost	10815.725 (8438.300, 14876.095)	112513.530 (99841.570, 16757.610)	-5.404	< 0.001
Western Medicine Cost	1822.425 (1094.910, 2592.600)	1990.820 (1330.660, 3242.570)	-3.771	< 0.001
Antibiotic Cost	312.995 (0.000, 592.383)	415.240 (137.200, 754.360)	-3.823	< 0.001
Length of Hospital Stay	5.000(4.000,7.000)	6.000(5.000,8.000)	-5.118	< 0.001
Hospitalizations for COPD in Past Year	1.000(1.000,1.000)	1.000(1.000,1.000)	-2.749	0.006
Historical Number Hospitalizations	1.000(1.00,2.000)	1.000(1.000,2.000)	-2.964	0.003

3.3. Comorbidity association rule analysis

In this study, COPD comorbidity association analysis, 15 chronic diseases were included in the exploration of association rule analysis based on Apriori algorithm by mining the combination of comorbid diseases of COPD patients through data, setting the preconditions as definitions of no more than 1, 2, and 3, respectively, with support > 1% and confidence > 50%, and the association rule mining identified 19 rules among COPD comorbidities. The specific reasons for these adjustments are explained in the supplementary file 3.

Comorbidity association analysis of the 2-way combination yielded 3 strong association rules (Table 4), namely “Arrhythmia and Other respiratory diseases”, “Diabetes and Hypertension”, and “Arthropathy and Hypertension” combination. Among these, “Arrhythmia and Other respiratory diseases” had the highest support (7.781%), and all three rules had a lift greater than 1, indicating that the presence of the former disease may increase the occurrence of the latter.

The comorbidity association analysis of the 3-way combination yielded a total of 13 strong association rules (Table 4), and the comorbidity association analysis of the 4-way combination yielded a total of 3 strong association rules (Table 5). The network diagram (Fig. 4) visualizes the association rules between COPD comorbidities, with arrows indicating the path of the comorbidities, the size of the circle representing the degree of support and the color of the circle representing the degree of lift. The graph shows that hypertension and cerebrovascular diseases have the highest degree of lift associated with heart diseases, indicating that if a COPD patient has both hypertension and cerebrovascular diseases, the likelihood of heart diseases is high. The size of the circles shows that the “Arrhythmia and Other respiratory diseases” and “Diabetes and Hypertension” rules are more strongly supported, indicating that COPD patients are more likely to have both arrhythmia and other respiratory diseases. More likely to have both diabetes and hypertension.

Table 4

Comorbidity association rules for 2-way combination
Left-Hand Side	Right-Hand Side	support	confidence	lift
Arrhythmia	Other respiratory diseases	7.781	51.692	1.122
Diabetes	Hypertension	6.068	52.4	1.628
Arthropathy	Hypertension	1.297	54.902	1.706

Table 5

Comorbidity association rules for 3-way combination
Left-Hand Side	Right-Hand Side	support	confidence	lift
Heart diseases, Arrhythmia	Other respiratory diseases	3.242	58.824	1.276
Hypertension, Arrhythmia	Other respiratory diseases	2.918	50.400	1.094
Diabetes, Other respiratory diseases	Hypertension	2.825	53.043	1.648
Heart diseases, Liver diseases	Hypertension	2.686	53.211	1.653
Diabetes, Heart diseases	Hypertension	2.501	62.791	1.951
Diabetes, Heart diseases	Other respiratory diseases	2.131	53.488	1.161
Heart diseases, Kidney diseases	Hypertension	1.760	61.290	1.904
Heart diseases, Cerebrovascular diseases	Hypertension	1.436	59.615	1.852
Hypertension, Cerebrovascular diseases	Heart diseases	1.436	53.448	2.564
Diabetes, Arrhythmia	Other respiratory diseases	1.390	63.83	1.385
Heart diseases, Cerebrovascular diseases	Other respiratory diseases	1.204	50.000	1.085
Diabetes, Arrhythmia	Hypertension	1.158	53.191	1.652
Heart diseases, Digestive diseases	Hypertension	1.019	61.111	1.898

Table 6

Comorbidity association rules for 4-way combination
Left-Hand Side	Right-Hand Side	support	confidence	lift
Hypertension, Heart diseases, Arrhythmia	Other respiratory diseases	1.390	53.571	1.162
Diabetes, Heart diseases, Other respiratory diseases	Hypertension	1.297	60.870	1.891
Hypertension, Diabetes, Heart diseases	Other respiratory diseases	1.297	51.852	1.125

3.4. Logistic regression of healthcare resource utilization in comorbidities.

Univariate and multivariate logistic regression analyses were performed using association rules: LOS and antimicrobial drug costs were used as dichotomous variables. Among the comorbidity combinations in the association rule, “Diabetes, Other respiratory diseases and Hypertension”, “Diabetes, Heart disease, Other respiratory diseases and Hypertension” was associated with LOS longer than or equal to 7 days. After adjustment for covariates including sex, age, education level, comorbidity numbers, BMI, ADL, and smoking, the adjusted OR (aOR) was still greater than 1 with a p-value less than 0.05(Table 7). Also, when median antimicrobial cost was used as a dichotomous variable, the combinations of “Diabetes and Hypertension”, “Diabetes, Other respiratory diseases and Hypertension”, “Diabetes, Heart disease, and Other respiratory diseases” had relatively higher antimicrobial drug costs, and after adjusting for covariates, their aOR was still greater than 1, with a p-value of less than 0.05(Table 8).

Table 7

Comorbidity Combinations and Length of Hospital Stay.
Comorbidity Combination	Percentage(%)	Percentage with Hospital Stay ≥ 7 Days	OR	aOR	95% CI	P-Value
Diabetes, Other respiratory diseases and Hypertension	2.83	4.76	3.93	2.30	1.18–4.46	0.014
Diabetes, Heart disease, Other respiratory diseases and Hypertension	1.30	2.38	5.80	3.03	1.03–8.88	0.044

Table 8

Comorbidity Combinations and Antibiotic Costs.
Comorbidity Combination	Percentage(%)	Percentage with Antibiotic Costs ≥ 400 RMB	OR	aOR	95% CI	P-Value
Diabetes and Hypertension	6.07	8.13	2.13	1.78	1.17–2.73	0.008
Diabetes, Other respiratory diseases and Hypertension	2.83	4.34	3.44	2.63	1.31–5.31	0.007
Diabetes, Heart disease, and Other respiratory diseases	2.13	3.32	3.67	2.55	1.15–5.65	0.021

4.1 Prevalence and Impact of Comorbidities in COPD Patients.

In the context of global population ageing, there has been an almost universal increase in life expectancy across countries and regions. However, ageing is associated with the incidence of chronic diseases[23], resulting in an increasing number of patients with chronic non-communicable diseases and a rising prevalence of comorbidities[24]. In China, with the ageing of the population and changes in the disease spectrum, comorbidity is becoming more common among the elderly; according to the China Health and Aged Care Tracking Survey (CHARLS), the prevalence of co-morbidity among the elderly over 45 years of age is 48.31%, while it rises to 73.86% among the elderly over 75 years of age[25]. With more than 400 million patients worldwide, COPD has become the third leading cause of death from chronic diseases, with 3.2 million patients dying from the disease each year[26]. Patients with COPD are usually accompanied by a variety of comorbidities such as cardiovascular disease, metabolic disorders, psychological problems and other disorders, which significantly affect the prognosis of the disease and the quality of life of the patients[27]. The prevalence of comorbidities in people with COPD varies according to the methodology of the study and the definition of comorbidities, but the conclusion is the same: people with COPD usually have multiple comorbidities. The present study found that the majority of the COPD population had at least one comorbidity, with a prevalence of 83.7%, a figure similar to the 84% reported in previous studies[28]. The most common comorbidities are other respiratory diseases, including asthma, bronchiectasis, pulmonary hypertension, pulmonary heart disease and respiratory failure, followed by hypertension. Studies have shown that asthma and COPD are often co-diagnosed in older patients[29] and that the rate of comorbidities is higher in patients with COPD than in those with asthma or COPD alone[30–32]. In addition, the prevalence of pulmonary hypertension in COPD patients is as high as 40%[33], and it has been reported that approximately 30% of COPD patients suffer from bronchiectasis[34]. The results of the present study are consistent with these studies, showing a high prevalence of other respiratory comorbidities in COPD patients, and showing a relatively high prevalence of hypertension, which is consistent with the results of other studies[35, 36].

4.2 Comorbidity in COPD Patients: Demographic, Clinical Characteristics, and Healthcare Resource Utilization.

This study found that the prevalence of comorbidities increased with age. Age-related changes in the lungs are associated with the onset of many comorbidities, including neurological, cardiovascular, psychiatric and gastrointestinal diseases, muscle wasting and osteoporosis[37]. This trend may be related to age-related physiological changes such as reduced immune function, impaired organ function and exacerbation of the chronic inflammatory state. Studies have shown that older COPD patients are more likely to develop cardiovascular disease and have more severe systemic inflammation and endothelial dysfunction than younger patients[38]. In addition, advancing age is usually associated with a greater accumulation of lifestyle risk factors, such as a history of smoking, which can increase the risk of comorbidities[39]. Therefore, integrated management and multidisciplinary care are particularly important in elderly patients with COPD to minimize the impact of comorbidities. This study also found that the prevalence of comorbidities was strongly associated with BMI. The data showed a significant increase in the prevalence of comorbidities among those with a low BMI (< 18.5 kg/m²) or high BMI (≥ 30 kg/m²). Obese patients have an increased risk of cardiovascular disease due to abnormal levels of several bioactive adipokines (e.g. leptin, lipocalin, etc.) released from adipose tissue, leading to an increased inflammatory response and oxidative stress[40]. Divo et al. also found in their study that the number of comorbidities increased as BMI increased[41]. This further emphasizes the importance of maintaining an appropriate body weight in patients with COPD. In clinical management, special attention should be given to weight management in patients with COPD to control weight through nutritional interventions and lifestyle changes to minimize comorbidities. Also, this study showed that COPD patients with comorbidities were significantly less able to care for themselves than patients without comorbidities. This finding is consistent with previous studies that the presence of comorbidities significantly reduces health-related quality of life in COPD patients[42]. This may be due to comorbidities that increase the burden of disease and lead to a reduction in patients' functional status and quality of life. Comorbidities such as cardiovascular disease, diabetes and osteoarthritis can limit patients' mobility and increase fatigue and pain, affecting their ability to care for themselves in daily life[43]. Therefore, there is a clinical need for a comprehensive functional assessment and rehabilitation program for these patients to improve their quality of life.

This study also found some differences in the clinical characteristics of COPD patients with and without comorbidities, in which NR, PCO2, Fbg, DD, BG, pro-BNP, PCT and TG were higher in the patients with comorbidities than in the group without comorbidities. The levels of Hb, PLT, pH and Alb were lower in the patients with comorbidities than in the group without comorbidities. The elevated NR may be associated with an increased chronic inflammatory response in COPD patients, which is consistent with previous studies[44]. Elevated PCO2 levels in COPD patients with comorbidity usually indicate further impairment of respiratory function and an increased risk of chronic respiratory failure. Elevated Fbg and DD may reflect the chronic low-grade inflammation and hypercoagulable state present in COPD patients, which is associated with an increased risk of COPD-related cardiovascular disease[45]. In addition, elevated (random) BG levels may be associated with insulin resistance or abnormal glucose metabolism in patients with COPD, suggesting that attention should be paid to the risk of diabetes in these patients[46]. Elevated pro-BNP and calcitonin are indicative of impaired cardiac function and the presence of underlying infection, and changes in these markers are usually seen in patients with comorbid cardiopulmonary disease. Elevated TG levels may be associated with metabolic syndrome and dyslipidemia, suggesting the need for improved cardiovascular risk management in patients with COPD[47]. Reduced Hb and Alb levels may reflect poor nutritional status in patients with comorbidity or chronic disease depletion. In contrast, the decrease in blood gas PH suggests a disturbance in acid-base balance, which may be associated with acute exacerbation of COPD or chronic respiratory failure with comorbidities. The above analysis shows that COPD patients with comorbidities are associated with complex changes in clinical features, requiring comprehensive assessment and individualized treatment in clinical management to improve prognosis.

COPD patients with comorbidities had higher total costs of the current hospitalization, western medicine costs, antimicrobial costs, LOS, number of hospitalizations for COPD in the past 1 year, and number of historical hospitalizations than the group without comorbidities. This may be since COPD patients with comorbidities require more complex treatment regimens and more healthcare resources, resulting in increased costs and LOS[48–50]. In addition, patients with comorbidities are more prone to acute exacerbations, increasing the frequency of hospitalization[51, 52]. Therefore, individualized treatment planning and enhanced management of comorbidities are needed to reduce the healthcare resource utilization on patients and improve treatment outcomes.

4.3 Comorbidity association rules and healthcare resource utilization

In this study, common comorbidity patterns of COPD were revealed by association rule analysis of comorbidity combinations of COPD patients. The results of association rules showed that COPD patients had the highest number of association rules related to other respiratory diseases and hypertension, and a variety of diseases such as arrhythmia, diabetes, arthropathy, heart diseases, liver diseases, kidney diseases, cerebrovascular diseases, etc. were associated with other respiratory diseases and hypertension, suggesting that more attention should be paid to the risk of other respiratory diseases and hypertension in patients with COPD while suffering from the aforementioned diseases and other diseases.

It was also found that certain combinations of comorbidities were associated with longer hospital stays and higher antimicrobial drug costs, and that longer hospital stays may reflect the complexity and severity of these combinations of comorbidities, such as diabetes, other respiratory diseases, hypertension and heart disease, resulting in the need for more medical resources and care[35]. In addition, the high cost of antimicrobials may reflect the fact that patients with these comorbidities are more susceptible to infections or have infections that are more difficult to treat and require more potent or longer-term antimicrobial therapy. These findings highlight the significant impact of comorbidities in COPD patients, particularly the combination of diabetes, hypertension, heart disease and other respiratory conditions. Longer hospital stays and higher antimicrobial drug costs not only increase the financial burden on patients, but also place greater demands on the healthcare system. To effectively manage these high-risk patients, healthcare providers need to develop comprehensive, multidisciplinary treatment plans that focus on early intervention and prevention to reduce length of stay and antimicrobial use.

4.4 Research innovations and limitations

The innovation of this study is that the association analysis of comorbidities using the association rule mining method identified diseases with a strong association in the pattern of COPD comorbidities and found that some of the associated diseases have a significant impact on the number of days of hospitalization and the cost of antimicrobial drugs in the burden of treatment. The shortcomings of this study are that it is based on a cross-sectional study, the samples are all from hospitals and the sample size is small, so it lacks representativeness, so the sample size can be increased and multi-centre studies can be conducted in future studies to increase the representativeness of the samples; also, there is no uniform standard for setting the parameters of the association rules for chronic diseases at home and abroad. Therefore, the parameters in this study were set by referring to the relevant literature and continuously adjusting the parameters to achieve satisfactory mining results, and the setting of parameter thresholds needs to be further explored.

In summary, this study, we analyzed the comorbidity patterns of 2159 COPD patients using association rule mining methods and found that certain combinations of comorbidities (e.g. diabetes mellitus, other respiratory diseases and hypertension) significantly increased the length of hospital stay and antimicrobial drug costs. These results suggest that multiple comorbidities have an important impact on healthcare resource use in COPD patients. Future studies should further investigate the specific mechanisms of these comorbidity combinations and develop optimal management strategies for high-risk comorbidity combinations, including multidisciplinary collaboration, precision medicine, patient education and policy support, to improve patient prognosis and reduce healthcare costs.

COPD Chronic Obstructive Pulmonary Disease

ICD-10 International Classification of Disease, 10th edition

LOS Length of Stay

ARM Association Rule Mining

BMI Body Mass Index

ADL Autonomy for Daily Living

aOR Adjusted Odds Ratio

CI Confidence Interval

NR Neutrophil Ratio

PCO2 Partial Pressure of Carbon Dioxide

Fbg Fibrinogen

DD D-Dimer

BG Blood Glucose

pro-BNP Pro-Brain Natriuretic Peptide

PCT Procalcitonin

Hb Hemoglobin

PLT Platelet Count

pH Blood Gas pH

PO2 Partial Pressure of Oxygen

ALT Alanine Aminotransferase

AST Aspartate Aminotransferase

TG Triglyceride

TC Total Cholesterol

HDL High-Density Lipoprotein

LDL. Low-Density Lipoprotein

ESR Erythrocyte Sedimentation Rate

Ethics approval and consent to participate

This study is a retrospective study, and the data used were obtained from previously conducted clinical projects. All data have been de-identified to protect patient privacy. The First Affiliated Hospital of Guangzhou Medical University has implemented a general informed consent procedure, and all inpatients have signed a general informed consent form allowing their clinical data to be used for research purposes. The study protocol has been reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Guangzhou Medical University (Approval Number: ES-2024-K085-01).

Clinical trial number

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

This research was supported by the Guangdong Medical Science and Technology Research Fund (Project No. C2022101).

Authors' contributions

Zhuozhuo Cheng: Participated in the writing of the manuscript.

Rui Zhang: Conducted data analysis.

Haofeng Xu: Participated in the interpretation of the data.

Junting Huang: Contributed to the conception and design of the study.

Zijing Liang: Participated in the revision of the manuscript.

Ping Yan: Supervised the overall work and approved the final version.

All authors participated in the writing of the manuscript and approved the final version.

Acknowledgements

We extend our gratitude to all the pioneers in respiratory research. Your groundbreaking work and invaluable support have been crucial to the success of this study.

Agustí A, Celli BR, Criner GJ, Halpin D, Anzueto A, Barnes P, Bourbeau J, Han MK, Martinez FJ, Montes de Oca M, et al: Global Initiative for Chronic Obstructive Lung Disease 2023 Report: GOLD Executive Summary. Eur Respir J 2023, 61.
Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, Abraham J, Adair T, Aggarwal R, Ahn SY, et al: Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012, 380:2095-2128.
Singh D, Agusti A, Anzueto A, Barnes PJ, Bourbeau J, Celli BR, Criner GJ, Frith P, Halpin DMG, Han M, et al: Global Strategy for the Diagnosis, Management, and Prevention of Chronic Obstructive Lung Disease: the GOLD science committee report 2019. Eur Respir J 2019, 53.
Zhong N, Wang C, Yao W, Chen P, Kang J, Huang S, Chen B, Wang C, Ni D, Zhou Y, et al: Prevalence of chronic obstructive pulmonary disease in China: a large, population-based survey. Am J Respir Crit Care Med 2007, 176:753-760.
Leal JR, Laupland KB: Validity of ascertainment of co-morbid illness using administrative databases: a systematic review. Clin Microbiol Infect 2010, 16:715-721.
National Guideline C: National Institute for Health and Care Excellence: Guidelines. In Multimorbidity: Assessment, Prioritisation and Management of Care for People with Commonly Occurring Multimorbidity. London: National Institute for Health and Care Excellence (NICE) Copyright © National Institute for Health and Care Excellence, 2016.; 2016
Eroglu SA, Gunen H, Yakar HI, Yildiz E, Kavas M, Duman D: Influence of comorbidities in long-term survival of chronic obstructive pulmonary disease patients. J Thorac Dis 2019, 11:1379-1386.
Santos NCD, Miravitlles M, Camelier AA, Almeida VDC, Maciel R, Camelier FWR: Prevalence and Impact of Comorbidities in Individuals with Chronic Obstructive Pulmonary Disease: A Systematic Review. Tuberc Respir Dis (Seoul) 2022, 85:205-220.
Antonelli Incalzi R, Fuso L, De Rosa M, Forastiere F, Rapiti E, Nardecchia B, Pistelli R: Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease. Eur Respir J 1997, 10:2794-2800.
James BD, Greening NJ, Tracey N, Haldar P, Woltmann G, Free RC, Steiner MC, Evans RA, Ward TJ: Prognostication of co-morbidity clusters on hospitalisation and mortality in advanced COPD. Respir Med 2024, 222:107525.
Tran PB, Kazibwe J, Nikolaidis GF, Linnosmaa I, Rijken M, van Olmen J: Costs of multimorbidity: a systematic review and meta-analyses. BMC Med 2022, 20:234.
Huber MB, Wacker ME, Vogelmeier CF, Leidl R: Excess costs of comorbidities in chronic obstructive pulmonary disease: a systematic review. PLoS One 2015, 10:e0123292.
Held FP, Blyth F, Gnjidic D, Hirani V, Naganathan V, Waite LM, Seibel MJ, Rollo J, Handelsman DJ, Cumming RG, Le Couteur DG: Association Rules Analysis of Comorbidity and Multimorbidity: The Concord Health and Aging in Men Project. J Gerontol A Biol Sci Med Sci 2016, 71:625-631.
Feng J, Mu XM, Ma LL, Wang W: Comorbidity Patterns of Older Lung Cancer Patients in Northeast China: An Association Rules Analysis Based on Electronic Medical Records. Int J Environ Res Public Health 2020, 17.
Ho VP, Schiltz NK, Reimer AP, Madigan EA, Koroukian SM: High-Risk Comorbidity Combinations in Older Patients Undergoing Emergency General Surgery. J Am Geriatr Soc 2019, 67:503-510.
Zhao P, Yoo I: Potentially modifiable risk factors for 30-day unplanned hospital readmission preventive intervention-A data mining and statistical analysis. Health Informatics J 2021, 27:1460458221995231.
Lee Y, Kim H, Jeong H, Noh Y: Patterns of Multimorbidity in Adults: An Association Rules Analysis Using the Korea Health Panel. Int J Environ Res Public Health 2020, 17.
Yu Z, Chen Y, Xia Q, Qu Q, Dai T: Identification of status quo and association rules for chronic comorbidity among Chinese middle-aged and older adults rural residents. Front Public Health 2023, 11:1186248.
Dencker K, Gottfries CG: Activities of daily living ratings of elderly people using Katz' ADL Index and the GBS-M scale. Scand J Caring Sci 1995, 9:35-40.
Hahsler M, Chelluboina S, Hornik K, Buchta C: The arules R-Package Ecosystem: Analyzing Interesting Patterns from Large Transaction Data Sets. J Mach Learn Res 2011, 12:2021-2025.
Wang CH, Lee TY, Hui KC, Chung MH: Mental disorders and medical comorbidities: Association rule mining approach. Perspect Psychiatr Care 2019, 55:517-526.
Liu Y, Wang L, Miao R, Ren H: A Data Mining Algorithm for Association Rules with Chronic Disease Constraints. Comput Intell Neurosci 2022, 2022:8526256.
Kuzuya M: Era of geriatric medical challenges: Multimorbidity among older patients. Geriatr Gerontol Int 2019, 19:699-704.
Palladino R, Pennino F, Finbarr M, Millett C, Triassi M: Multimorbidity And Health Outcomes In Older Adults In Ten European Health Systems, 2006-15. Health Aff (Millwood) 2019, 38:613-623.
Zhang L, Wei L, Fang Y: Spatial-temporal distribution patterns and influencing factors analysis of comorbidity prevalence of chronic diseases among middle-aged and elderly people in China: focusing on exposure to ambient fine particulate matter (PM(2.5)). BMC Public Health 2024, 24:550.
Adeloye D, Chua S, Lee C, Basquill C, Papana A, Theodoratou E, Nair H, Gasevic D, Sridhar D, Campbell H, et al: Global and regional estimates of COPD prevalence: Systematic review and meta-analysis. J Glob Health 2015, 5:020415.
Almagro P, Soler-Cataluña JJ, Huerta A, González-Segura D, Cosío BG: Impact of comorbidities in COPD clinical control criteria. The CLAVE study. BMC Pulm Med 2024, 24:6.
Schnell K, Weiss CO, Lee T, Krishnan JA, Leff B, Wolff JL, Boyd C: The prevalence of clinically-relevant comorbid conditions in patients with physician-diagnosed COPD: a cross-sectional study using data from NHANES 1999-2008. BMC Pulm Med 2012, 12:26.
Soriano JB, Visick GT, Muellerova H, Payvandi N, Hansell AL: Patterns of comorbidities in newly diagnosed COPD and asthma in primary care. Chest 2005, 128:2099-2107.
Miravitlles M, Soriano JB, Ancochea J, Muñoz L, Duran-Tauleria E, Sánchez G, Sobradillo V, García-Río F: Characterisation of the overlap COPD-asthma phenotype. Focus on physical activity and health status. Respir Med 2013, 107:1053-1060.
Blanchette CM, Gutierrez B, Ory C, Chang E, Akazawa M: Economic burden in direct costs of concomitant chronic obstructive pulmonary disease and asthma in a Medicare Advantage population. J Manag Care Pharm 2008, 14:176-185.
Rhee CK, Yoon HK, Yoo KH, Kim YS, Lee SW, Park YB, Lee JH, Kim Y, Kim K, Kim J, et al: Medical utilization and cost in patients with overlap syndrome of chronic obstructive pulmonary disease and asthma. Copd 2014, 11:163-170.
Chaouat A, Naeije R, Weitzenblum E: Pulmonary hypertension in COPD. Eur Respir J 2008, 32:1371-1385.
Huang JT, Cant E, Keir HR, Barton AK, Kuzmanova E, Shuttleworth M, Pollock J, Finch S, Polverino E, Bottier M, et al: Endotyping Chronic Obstructive Pulmonary Disease, Bronchiectasis, and the "Chronic Obstructive Pulmonary Disease-Bronchiectasis Association". Am J Respir Crit Care Med 2022, 206:417-426.
Mannino DM, Thorn D, Swensen A, Holguin F: Prevalence and outcomes of diabetes, hypertension and cardiovascular disease in COPD. Eur Respir J 2008, 32:962-969.
Curkendall SM, DeLuise C, Jones JK, Lanes S, Stang MR, Goehring E, Jr., She D: Cardiovascular disease in patients with chronic obstructive pulmonary disease, Saskatchewan Canada cardiovascular disease in COPD patients. Ann Epidemiol 2006, 16:63-70.
Cazzola M, Bettoncelli G, Sessa E, Cricelli C, Biscione G: Prevalence of comorbidities in patients with chronic obstructive pulmonary disease. Respiration 2010, 80:112-119.
Tudorache E, Fira-Mladinescu O, Traila D, Marc M, Rajnoveanu RM, Tofolean DE, Fildan AP: Endothelial dysfunction: The possible link between cardiovascular comorbidities and phenomenon of inflammaging from COPD. Medicine (Baltimore) 2022, 101:e30078.
Fabbri LM, Luppi F, Beghé B, Rabe KF: Complex chronic comorbidities of COPD. Eur Respir J 2008, 31:204-212.
Mafort TT, Rufino R, Costa CH, Lopes AJ: Obesity: systemic and pulmonary complications, biochemical abnormalities, and impairment of lung function. Multidiscip Respir Med 2016, 11:28.
Divo MJ, Cabrera C, Casanova C, Marin JM, Pinto-Plata VM, de-Torres JP, Zulueta J, Zagaceta J, Sanchez-Salcedo P, Berto J, et al: Comorbidity Distribution, Clinical Expression and Survival in COPD Patients with Different Body Mass Index. Chronic Obstr Pulm Dis 2014, 1:229-238.
Wacker ME, Jörres RA, Karch A, Koch A, Heinrich J, Karrasch S, Schulz H, Peters A, Gläser S, Ewert R, et al: Relative impact of COPD and comorbidities on generic health-related quality of life: a pooled analysis of the COSYCONET patient cohort and control subjects from the KORA and SHIP studies. Respir Res 2016, 17:81.
Putcha N, Puhan MA, Hansel NN, Drummond MB, Boyd CM: Impact of co-morbidities on self-rated health in self-reported COPD: an analysis of NHANES 2001-2008. Copd 2013, 10:324-332.
Barnes PJ: The cytokine network in chronic obstructive pulmonary disease. Am J Respir Cell Mol Biol 2009, 41:631-638.
Sin DD, Man SF: Why are patients with chronic obstructive pulmonary disease at increased risk of cardiovascular diseases? The potential role of systemic inflammation in chronic obstructive pulmonary disease. Circulation 2003, 107:1514-1519.
Lin CS, Liu CC, Yeh CC, Chang YC, Chung CL, Lane HL, Shih CC, Chen TL, Liao CC: Diabetes risks and outcomes in chronic obstructive pulmonary disease patients: Two nationwide population-based retrospective cohort studies. PLoS One 2017, 12:e0181815.
Xuan L, Han F, Gong L, Lv Y, Wan Z, Liu H, Zhang D, Jia Y, Yang S, Ren L, Liu L: Association between chronic obstructive pulmonary disease and serum lipid levels: a meta-analysis. Lipids Health Dis 2018, 17:263.
Doos L, Uttley J, Onyia I, Iqbal Z, Jones PW, Kadam UT: Mosaic segmentation, COPD and CHF multimorbidity and hospital admission costs: a clinical linkage study. J Public Health (Oxf) 2014, 36:317-324.
Guo Ie H, Tang CH, Sheu ML, Liu HY, Lu N, Tsai TY, Chen BL, Huang KC: Evaluation of risk adjustment performance of diagnosis-based and medication-based comorbidity indices in patients with chronic obstructive pulmonary disease. PLoS One 2022, 17:e0270468.
Chen W, FitzGerald JM, Sin DD, Sadatsafavi M: Excess economic burden of comorbidities in COPD: a 15-year population-based study. Eur Respir J 2017, 50.
Kim Y, Kim YJ, Kang YM, Cho WK: Exploring the impact of number and type of comorbidities on the risk of severe COPD exacerbations in Korean Population: a Nationwide Cohort Study. BMC Pulm Med 2021, 21:151.
Seo H, Sim YS, Min KH, Lee JH, Kim BK, Oh YM, Ra SW, Kim TH, Hwang YI, Park JW: The Relationship Between Comorbidities and Microbiologic Findings in Patients with Acute Exacerbation of Chronic Obstructive Pulmonary Disease.Int J Chron Obstruct Pulmon Dis 2022, 17:855-867.

No competing interests reported.

Supplementaryfiles.zip
Supplementary files Supplementary file 1 Title of data: The number of comorbidities in COPD patients. Description of data: This file details the distribution of the number of comorbidities among COPD patients. The table includes the number of cases and the corresponding percentage for patients with 0, 1, 2, 3, and 4 or more comorbidities. Supplementary file 2 Title of data: Distribution of Comorbidities by Gender, Smoking History, and Educational Level. Description of data: This file details the distribution of comorbidities among COPD patients based on gender, smoking history, and educational level. The table includes the number of cases, the number of comorbidities, the comorbidity rate, chi-square (X²) values, and p-values for each characteristic category. Supplementary file 3 Title of data: Explanation of Support and Confidence Threshold Adjustments in Association Rule Mining Description of data: This file provides a detailed explanation of the adjustments made to the support and confidence thresholds used in the association rule mining process. The dataset for this study contains 2159 records, and a minimum support level of 1% was chosen to ensure that comorbidity combinations occurred in at least 1% of cases, making them statistically significant and practically meaningful. A minimum confidence level of 50% was chosen to ensure high reliability and clinical importance of the identified comorbidity associations. The file also discusses the rationale behind these thresholds and the resultant 19 strong association rules identified under these conditions. Supplementary file 4 Title of data: Raw Data Table Description of data: This file includes the raw data used in the study, containing all initial measurements and observations before any data processing or analysis. Supplementary file 5 Title of data: Post-Exclusion Data Table Description of data: This file contains the data table after excluding certain participants based on specific criteria. Supplementary file 6 Title of data: Table after Multiple Interpolation Description of data: This file includes the data table after applying multiple interpolation techniques to handle missing values.

Download PDF

Version 1

posted

You are reading this latest preprint version

Comorbidity Patterns and Healthcare Resource Utilization in COPD Patients: A Retrospective Analysis Using Association Rules

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

1. Introduction

2. Material and methods

2.1 Selected objects

2.2 Methods

2.3. Statistical analyses

3. Results

3.1. Comorbidities in COPD patients

3.2. Comorbidity in COPD Patients: Demographics, Clinical Characteristics, and Healthcare Resource Utilization

3.3. Comorbidity association rule analysis

3.4. Logistic regression of healthcare resource utilization in comorbidities.

4. Discussion

4.1 Prevalence and Impact of Comorbidities in COPD Patients.

4.2 Comorbidity in COPD Patients: Demographic, Clinical Characteristics, and Healthcare Resource Utilization.

4.3 Comorbidity association rules and healthcare resource utilization

4.4 Research innovations and limitations

5. Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1