Impact of multidisciplinary team care on reducing utilization of emergency department visits for patients with lung cancer.
To improve the quality of care, multidisciplinary team (MDT) care was implemented in Taiwan. This study examined the relationship between MDT care and emergency department (ED) visits for lung cancer patients.
A retrospective cohort study with MDT care participants and matched a double number of control group of non-participants was followed.
In this study, 22,817 patients with newly diagnosed lung cancer were recruited from 2005 to 2007 in Taiwan. Matching based on the propensity of receiving MDT care was used. A total of 8172 patients were observed in this study. A c2, ANOVA, logistic regression, and Poisson regression were used to elucidate the effects of MDT care.
The lung cancer patients participating in MDT had lower risk to visit an ED (OR = 0.89; 95% CI, 0.80-0.98), and the incidence rate ratio decreased by 11% (95% CI, —0.15 to –0.07). Gender, monthly salary, urbanization of the residence area, comorbid conditions, catastrophic illness/injury, treatment method, number of outpatient visits, length of stay,
hospital ownership, level of hospital, and the age of the patient’s physician were all significantly related to the frequency of ED visits (P <.05).
The frequency of ED visits of patients with MDT care was lower than that of those without it. The patients with MDT received enhanced care.
Am J Manag Care. 2014;20(8):e353-e364
Worldwide, lung cancer is the most common cancer
in terms of both incidence and mortality. In
2012, there were 1.8 million new cases and 1.59
million deaths caused by lung cancer.1 Since 1998, lung cancer
has been the leading cause of cancer death in Taiwan
and worldwide.2 Although progress has been made in the
diagnosis and treatment of lung cancer, both the incidence
and mortality rate of lung cancer have increased in recent
years. Formal multidisciplinary team (MDT) care has been
widely promoted worldwide to improve coordination, communication,
and decision making in cancer management.3,4
The Taiwan bureau of National Health Insurance has implemented
“multidisciplinary team care for cancer patients”
since April 2003 to enhance the quality of cancer diagnosis
and treatment. The bureau emphasizes an MDT approach
that provides a complete cancer treatment plan for patients.
MDT care goes beyond such conventional treatments for
lung cancer as surgical excision, radiation therapy, and chemotherapy.
The MDT members can include related clinical
physicians, nursing staff, a psychological consultant, a social
worker, and a case manager to discuss a dedicated treatment
plan and to integrate all treatments and care. Therefore,
the patient with MDT care should benefit from stable and
continuous care that includes regular outpatient visits and
inpatient treatment, all arranged and coordinated through
the case manager. The Bureau of National Health Insurance
(NHI) paid additional fees to physicians to make MDT care
The MDT approach has been used for years in numerous
countries. Studies from the United States, Germany, the
United Kingdom, and Australia have demonstrated that an
MDT that integrates surgeons, tumor physicians, radiology
physicians, psychologists, physiatrists, and dietitians can improve
the quality of life for cancer patients,5 lower the cost
of healthcare,6 and increase the satisfaction of treatment for
patients,6 efficiency of treatment,4,5,7,8 and survival rate.9,10 In
Taiwan, Wang et al found the relative risk of death was lower
for oral cavity cancer in MDT care participants.4 Chen et
al found that chronic kidney disease patients who participated
in MDT had a better survival rate than nonparticipants
have initiate renal replacement therapy instead of
after MDT intervention.11
Hospital emergency departments (EDs) are typically designed
to manage emergent or unexpected situations and
are generally crowded and busy.12 As detailed in 1 study,
in North Carolina in 2008, 0.9% of ED visits were cancer
related; 7.7% of the state’s cancer survivors visited the ED;
and each ED visitor received 1.4 ED services per year on
average.13 In Taiwan, in 2012, 1.9% of ED visits were cancer
related.14 The reasons cancer patients seek ED services
include pain, dyspnea, nausea, and vomiting, among others.
13,15 Among cancer patients, lung cancer patients are,
as a group, the likeliest of all to seek ED services.13 The
main reasons that prompt the visits of lung cancer patient
to EDs include respiratory symptoms, fever, neurological/
psychiatric issues, and digestive complaints.16
The high percentage of ED visits for cancer patients
has been recognized as an indicator that end-of-life cancer
care is of less-than-ideal quality. Better care, it is believed,
could help cancer patients avoid at least some of the urgent
medical problems that necessitate ED visits.
Most studies on lung cancer in MDT care have been
limited, sometimes because the sample size was small and
sometimes because the focus was narrowly on survival.
Few studies have examined the changes in the utilization
of ED services after patients have become involved in
MDT care. Therefore, this study investigates the influence
of the participation and nonparticipation of patients
in MDT care on the utilization of ED visits.
This retrospective and longitudinal cohort study analyzed
the Cancer Dataset in the National Health Insurance
Research Database (from 1997 to 2008) published
by the Taiwan National Health Research Institutes and
provided by the Bureau of National Health Insurance
of Taiwan. The database included
the medical records of all individuals
insured by National Health Insurance
in Taiwan. At the end of 2009,
a total of 23,026,000 people were insured,
constituting 99.59% of Taiwan’s
population. From these data we could
extrapolate the utilization of health
services by the lung cancer patients in
Taiwan. The Bureau of National Health Insurance implemented
MDT care in April 2003, and this study used the
years of suffering from lung cancer diagnosis for cancer
treatments (the major
International Classification of Diseases,
Ninth Revision, Clinical Modification
1. Participants. According to the database, 8172 lung
cancer patients in Taiwan were newly diagnosed between
2005 and 2007. New patients defines a first-time lung cancer
diagnosis based on
code 162.x, and diagnosis
following a year of chemotherapy, radiation treatment,
surgery, or other cancer treatment. We followed up with
each patient within 1 year after lung cancer diagnosis. We
excluded those patients who had not received any treatment
since their diagnosis, and we also excluded the patients
who died within 1 month after diagnosis (N = 2050
patients). Since the physicians could decide whether the
patients participated in the MDT care or not, the MDT
participants might exhibitbias selection. Therefore, to reduce
the bias selection from this population-based data,
this study used a propensity score matching to evaluate
the likelihood of participating in MDT care for each patient
and so selected double patients as the control group.
In beginning, the study selected 22,817 lung cancer patients
which were 2736 patients who joined an MDT care
and 20,081 patients who did not. After propensity score
matching, there were 2724 MDT patients and double patients
in non-MDT groups.
2. Variable Description and Defintion. Independent
variables included patients’ gender, age, monthly salary,
urbanization of their area of residence, Charlson Comorbidity
Index (CCI) score, catastrophic illness number of
cancer outpatient visits, number of hospitalizations, and
participation status in MDT care, as well as certain characteristics
of their hospital and physician. Dependent
variables included whether the patients used ED service
or not, and the frequency of ED use in the 12 months following
diagnosis, excluding the utilization of the ED due
Study participants were categorized by gender (male,
female); age (divided into 7 groups from less than 24 years
to 75 years and older); monthly salary (divided into 8
groups); urbanization of residence area (noted on a scale
of 1 to 7,17 with 1 being the most urban and 7 the least);
CCI score (the extent of comorbidity was adapted by
Deyo);18 catastrophic illness (yes or no); number of cancer
outpatient visits (divided into 7 groups, from 0 visits to 26
and above); length of stay of hospitalization due to lung
cancer within 1 year after cancer diagnosis (LOS; divided
into 5 groups); ownership of the hospital visited most frequently
(public or non-public); level of that hospital (first
level, second level, third level, and clinic); and characteristics
(gender, age) of the patient’s physician.
3. The study used descriptive statistics, x2 test, test,
and ANOVA to analyze the characteristics of demography,
health status, disease treatment method, utilization
of healthcare within 1 year, hospital and physicians’
characteristics, and the differences in ED visits between
those who did and did not participate in MDT care. Logistic
regression and Poisson regression were used to analyze
the inference factors of probability and frequency of
ED visits between the patients who did and did not participate
in MDT care. All analyses were performed using
the sets statistical package version 9.2.
and show the characteristic distribution
before and following PSM of the sample. Following PSM,
8172 subjects were selected for this study. The characteristics
included the gender, age, monthly salary, urbanization
of residence area, CCI score, catastrophic illness, and
treatment methods of MDT and non-MDT participants.
The differences between the 2 groups on all these variable
measures were not statistically significant (
shows the top 20 causes of lung cancer patients’
ED visits. The primary cause was fever for both
the MDT participants (25.46%) and non-MDT partici
pants (23.97%), followed by dyspnea and respiratory abnormalities;
chest pain; abdominal pain; and dizziness
and fainting. These 5 causes account for 64% of the
reasons lung cancer patients visited the ED, and there
were no significant differences in this regard between the
MDT and non-MDT patients. Fewer MDT participants
(7.94%) visited EDs because of lung cancer-related chest
pains than non-MDT participants (10.2%). Furthermore,
fewer MDT participants (1.06%) visited EDs because
of coughing than did non-MDT participants (2.24%).
Among the top 20 causes for the MDT participants’
ED care visits were general symptoms (2.88%), convulsions
(1.7%), and other nervous and musculoskeletal
system symptoms (0.59%). For non-MDT participants,
ED visit reasons included alteration of consciousness
(1.75%), shock without mention of trauma (1.02%), and
general symptoms (0.56%).
of the ED care uses of MDT participants and non-MDT
participants who had different demographic variables
and treatment characteristics. Before controlling for
other factors, the results of the c2 test showed that MDT
participants had less chance to use ED services than
non-MDT paticipants, but not significantly. However,
the number of ED care visits in MDT participants was
significantly less than non-MDT participants. Patients
who had higher percentages of ED service uses tended
to be male; have monthly salaries lower than NT$57,800
(New Taiwan dollars); live in less urbanized areas; have
CCI scores greater than 7; have a greater likelihood to
have suffered a catastrophic illness; have made more
cancer outpatient visits; have had a longer LOS. Patients
who made more ED visits also tended to be receiving the
treatments of radiology therapy; or radiology therapy +
chemotherapy; or surgery + radiology therapy + chemotherapy.
In addition, patients who are treated by physicians
aged 44 years or less have significantly higher
percentages of ED service use (
Patients in our study with the following characteristics
had greater mean numbers of ED service uses: non-MDT
participants (mean = 1.53 ± standard deviation [SD] 2.63);
those between 25 and 34 years of age (mean = 2.37 ± SD
4.25); those with a, CCI score of 12 or greater (mean =
2.67 ± SD 3.01); residing at levels 4 and 5 in urbanization
of residence area (mean = 1.66 ± SD 3.21); those with
catastrophic injury or illness beyond lung cancer (mean =
1.51 ± SD 2.54); greater number of cancer outpatient visits
(mean = 1.66 ± SD 2.93), longer LO; and receiving surgery
+ radiology therapy + chemotherapy (mean = 1.72 ± SD
2.98). Patients who are treated at public hospitals have
greater mean numbers of ED service uses (mean = 1.57
± SD 2.39), and patients who are treated by physicians
55 years or older have the smallest mean number of ED
service uses (mean = 1.14 ± 1.88), a statistically significant
difference in comparison to patients treated by younger
shows the results of the logistic regression and
Poisson regression that were performed on the variables.
The results of the logistic regression show that MDT participants
have a lower likelihood of seeking ED services
(OR = 0.89; 95% CI, 0.80-0.98), which differs significantly
from non-MDT participants (
= .022). Female patients are
less likely to visit EDs compared with male patients (OR
= 0.73; 95% CI, 0.66-0.81); patients with monthly salaries
either between NT$57,801 and NT$72,800, or NT$72,800
or more, are less likely to visit EDs compared with patients
with premium-based monthly salaries of NT$17,280
or less (OR = 0.34; 95% CI, 0.21-0.54; OR = 0.52; 95% CI,
0.30-0.91); patients with CCI scores between 4 and 6, between
7 and 9, between 10 and 12, and 12 or greater are
more likely to visit EDs compared with patients with CCI
scores of 3 or less (OR = 1.60; 95% CI, 1.28-2.00; OR =
2.62; 95% CI, 2.13-3.22; OR = 3.70; 95% CI, 3.02-4.52; OR
= 7.82; 95% CI, 5.31-11.54); and patients with catastrophic
illness are more likely to visit EDs compared with patients
without catastrophic illness (OR = 2.94; 95% CI, 2.01-4.31).
Patients receiving surgery + radiology therapy, surgery +
chemotherapy, radiology therapy + chemotherapy, and
surgery + radiology therapy + chemotherapy are more
likely to visit EDs compared with patients receiving only
surgeries (OR = 1.89; 95% CI, 1.54-2.32; OR = 1.90; 95% CI,
1.57-2.30, OR = 1.61; 95% CI, 1.17-2.22; OR = 2.71; 95%
CI, 2.25-3.28). Patients with more than 1 outpatient visit
and treated in a district hospital, rather than a first-level
hospital, are more likely to visit the ED. Patients with longer
LOS and who are treated by physicians over 35 years
of age are less likely to visit the ED. These differences are
all statistically significant (
The results of the Poisson regression showed that after
controlling for the other variables, the MDT participants
visited EDs fewer times compared with non-MDT participants
(β = -0.09; 95% CI, —0.13 to 0.05). Female patients
visited EDs fewer times compared with male patients
(β = -0.18; 95% CI, —0.22 to 0.14). Patients who were insured
dependents or who had monthly salaries between
NT$17,281 and NT$22,800 visited EDs more times compared
with patients with monthly salaries of NT$17,280
or less (β = 0.10; 95% CI, 0.04-0.16; and β = 0.06; 95% CI,
0.00-0.11), whereas patients with premium-based monthly
salaries between NT$36,301 and NT$45,800, between
NT$57,801 and NT$72,800, and NT$72,801 or greater,
visited EDs fewer times compared with patients having
monthly salaries of NT$17,280 or less (β = —0.17; 95% CI,
—0.28 to 0.06; β = -0.43; 95% CI, –0.64 to 0.22; β = –0.42;
95% CI, —0.68 to 0.17). Patients at levels 2 and 3 and 4 and
5 of urbanization of residence area visited EDs more times
compared with patients at level 1 (the most urban) of urbanization
of residence area (β = 0.06; 95% CI, 0.01-0.11;
β = 0.16; 95% CI, 0.13-0.22). Patients with CCI scores between
4 and 6, between 7 and 9, between 10 and 12, and of
12 or more visited EDs more times compared with patients
with CCI scores of 3 or less (β = 0.39; 95% CI, 0.27-0.50; β
= 0.65; 95% CI, 0.54-0.76; β = 0.95; 95% CI, 0.85-1.06; and
β = 1.32; 95% CI, 1.19-1.45). Patients with catastrophic illness
visited EDs more times compared with patients without
them (β = 0.63; 95% CI, 0.41-0.85). Patients receiving
radiology therapy, surgery + radiology therapy, surgery +
chemotherapy, radiology therapy + chemotherapy, and
surgery + radiology therapy + chemotherapy visited EDs
more times compared with patients receiving only surgeries
(β = 0.32; 95% CI, 0.18-0.47; β = 0.34; 95% CI, 0.24-0.43;
β = 0.52; 95% CI, 0.43-0.61; β = 0.26; 95% CI, 0.13-0.40; β
= 0.66; 95% CI, 0.57—0.74). Patients receiving treatment at
private hospitals visited EDs fewer times compared with
patients receiving treatments at public hospitals (β = —0.11;
95% CI, -0.15 to 0.06). Patients with more than 1 outpatient
visit and treated in a second-level hospital or thirdlevel
hospital, rather than a medical center, visited the ED
more times. Patients with longer LOS and treated by physicians
aged 45 years or more visited the ED fewer times.
These differences were all statistically significant (
Our results showed that MDT participants visited EDs
fewer times compared with non-MDT participants. Although
EDs can provide immediate care to relieve acute
symptoms, cancer patients cannot receive proper and
holistic care because EDs, which are frequently crowded
and busy, are not satisfactory environments for such
care.13 In learning that MDT participants use ED care less
frequently, indicating that MDT care reduces lung cancer
patients’ needs to use ED services, we can infer that MDT
care improves the quality of cancer patient care.
The most common cause of cancer patients’ ED visits
is fever (approximately 25%), which differs from the most
common cause—pain—identified in previous studies.13,15,19
This is possibly because various countries adopt different
standards in the timing and methods of prescribing anti-infectives
and antipyretic and analgesic medication. Further
analysis must be conducted to examine cancer patients’
medication dosage and uses. In the 2008 study by Mayer
et al,13 approximately 7.7% of North Carolina’s cancer patients
(having all kinds of cancer, and ranging from recently
diagnosed to diagnosed years earlier) sought ED care, whereas
57% of the lung cancer patients in this study sought ED
care. This is possibly because (a) this study examined the
selected cancer patients’ ED service uses for the first year after
they were diagnosed by physicians, whereas Mayer et al
investigated all cancer patients’ ED service uses during that
year, which resulted in the substantial difference; or (b) the
charge of ED service in Taiwan was inexpensive ($5 to $15
US) compared with that in the United States.
Male patients use ED services more frequently than
female patients, which is consistent with the results of
previous studies.13,20 Cancer patients with catastrophic illness
use ED services more frequently, which is consistent
with the findings in the study conducted by Liu et al16 in
2006. Patients with higher CCI scores have more severe
comorbidities and thus use medical services, such as the
ED, more frequently. Patients earning higher salaries are
less likely to use ED services and use their services less
frequently compared with patients earning lower salaries,
possibly because the former are more capable of maintaining
and enhancing the quality of the daily care they receive
and are, therefore, less likely to require ED care. Patients
residing in less urbanized areas are more likely to use ED
services and to use them more frequently compared with
patients residing in more urbanized areas, indicating differing
access to medical services based on the residential
area’s level of urbanization. Patients whose cancer is being
treated by surgery alone use ED services less frequently
compared with patients receiving other treatments, possibly
because they had relatively simple procedures that
caused few complications or because they might have been
diagnosed relatively early, with less of a chance to develop
issues needing ED attention. In addition, lung cancer patients
receiving only chemotherapy use ED services less
frequently, although the difference is not statistically significant.
Patients receiving treatments at private hospitals
and public hospitals do not significantly differ regarding
whether they seek ED care. However, patients receiving
ongoing treatments at private hospitals visited EDs fewer
times compared with patients receiving treatments at
public hospitals, possibly reflecting a gap between private
and public care quality. As for physicians’ characteristics,
gender has no effect on patients’ ED service uses. In contrast,
patients who are treated by senior physicians are
less likely to use ED services and use these services less
frequently. This could indicate that physicians with more
medical experience can reduce their patients’ urgent and
unpredictable needs regarding receiving medical services.
PSM was employed to reduce the selection bias resulting
from varied patients’ characteristics in MDT participation.
The matching was based on variables including
gender, age, monthly salary, urbanization of residence
area, CCI scores, catastrophic illness/injury, treatment
methods, whether the patient’s hospital was public or private,
and characteristics of the patient’s physician. However,
we did not include certain factors that may have
affected patients’ ED service uses, such as smoking and
drinking habits, cancer staging, and type of lung cancer.
These data types are difficult to obtain and constitute
a limitation of this study. The results only showed the
strong potential causation but not the direct evidence.
This study found that lung cancer patients who participate
in MDT care use ED services less frequently.
EDs cannot provide cancer patients with holistic care.
Therefore, cancer patients should be provided with
MDT care to reduce their need for ED services. In addition,
patients participating in MDT care are less likely
to have urgent and unpredictable needs for medical care,
indicating that MDT care enhances the management of
lung cancer patients.
This study found that the primary cause of lung cancer
patients’ ED visits is fever, which differs from the findings
of studies conducted in other countries. Future studies can
be conducted to analyze the dosages of anti-infection and
antipyretic and analgesic medications that physicians prescribe
to lung cancer patients to elucidate the causes of the
high frequency of fever in Taiwanese lung cancer patients.
The authors are grateful for financial support from China Medical
University and Asia University (grant numbers: CMU100-ASIA-10,
DOH102-TD-B-111-004) as well as the National Science Council (grant
number: NSC98-2410-H-468-015-MY2) in Taiwan. They also deeply appreciate
the provision of the dataset of cancer patients by the National
Health Research Institutes.
Author Affiliations: Department of Public Health and Department of
Health Services Administration, China Medical University, Taichong,
Taiwan, ROC (SMW); Department of Health Services Administration,
China Medical University, Taichung, Taiwan, ROC (YHW, KHH,
WCT); Department of Health Care Administration, Oriental Institute of
Technology, Taipei, Taiwan, ROC (SMW); and Department of Healthcare
Administration, Asia University, Taichung, Taiwan, ROC (PTK).
Funding Source: This study was funded by China Medical University
and Asia University (grant numbers: CMU100-ASIA-10, DOH102-
TD-B-111-004) and the National Science Council (grant number:
Author Disclosures: The authors report no relationship or financial
interest with any entity that would pose a conflict of interest with the
subject matter of this article.
Authorship Information: Concept and design (SMW, WCT); analysis
and interpretation of data (PTK, SMW, KHH, WCT, YHW); drafting of
the manuscript (SMW); critical revision of the manuscript for important
intellectual content (KHH, WCT); statistical analysis (PTK, YHW); funding
(KHH, WCT); administrative, technical, or logistic support (KHH);
and supervision (PTK, WCT).
Address correspondence to: Wen-Chen Tsai, DrPH, No. 91 Hsueh-Shih
Road, Taichung, Taiwan 40402, ROC. E-mail: firstname.lastname@example.org.
GLOBOCAN 2012: estimated cncerincidence,
1. World Health Organization.
mortality and prevalence worldwide in 2012, 2012. http://globocan.
2. Department of Health EY. The Couse of Death Statistics in 2011 May
25, 2012 ed. ROC (Taiwan); 2012.
3. Price A, Kerr G, Gregor A, et al. The impact of multidisciplinary
teams and site specialisation on the use of radiotherapy in elderly
people with non-small cell lung cancer (NSCLC) [abstract].
4. Wang YH, Kung PT, Tsai WC, Tai CJ, Liu SA, Tsai MH. Effects of multidisciplinary
care on the survival of patients with oral cavity cancer in
5. Boxer MM, Vinod SK, Shafiq J, Duggan KJ. Do multidisciplinary
team meetings make a difference in the management of lung cancer?
6. Gabel M, Hilton NE, Nathanson SD. Multidisciplinary breast cancer
clinics. do they work?
7. Coory M, Gkolia P, Yang IA, Bowman RV, Fong KM. Systematic
review of multidisciplinary teams in the management of lung cancer.
8. Forrest LM, McMillan DC, McArdle CS, Dunlop DJ. An evaluation of
the impact of a multidisciplinary team, in a single centre, on treatment
and survival in patients with inoperable non-small-cell lung cancer.
9. Westin T, Stalfors J. Tumour boards/multidisciplinary head and neck
cancer meetings: are they of value to patients, treating staff or a political
additional drain on healthcare resources?
Curr Opin Otolaryngol
Head Neck Surg.
10. Birchall M, Bailey D, King P; South West Cancer Intelligence Service
Head and Neck Tumour Panel. Effect of process standards on survival
of patients with head and neck cancer in the south and west of England.
Br J Cancer.
11. Chen YR, Yang Y, Wang SC, et al. Effectiveness of multidisciplinary
care for chronic kidney disease in Taiwan: a 3-year prospective cohort
Nephrol Dial Transplant.
12. Barbera L, Paszat L, Chartier C. Indicators of poor quality end-of-life
cancer care in Ontario.
J Palliat Care.
13. Mayer DK, Travers D, Wyss A, Leak A, Waller A. Why do patients
with cancer visit emergency departments? results of a 2008 population
study in North Carolina.
J Clin Oncol.
14. Ministry of Health and Welfare EY. The Medical Statictical Annual
Report in 2012. Jan 28, 2014 ed. ROC (Taiwan); 2014. http://www.mohw.
15. Barbera L, Taylor C, Dudgeon D. Why do patients with cancer visit
the emergency department near the end of life?
16. Gorham J, Ameye L, Berghmans T, Sculier JP, Meert AP. The lung
cancer patient at the emergency department: a three-year retrospective
17. Liu CY, Hung YT, Chuang YL, et al. Incorporating development stratification
of Taiwan townships into sampling design of large scale health
J Health Manag (Chin).
18. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index
for use with ICD-9-CM administrative databases.
J Clin Epidemiol.
19. Tsai SC, Liu, LN, Tang ST, Chen JC, Chen, ML. Cancer pain as the
presenting problem in emergency departments: incidence and related
Support Care Cancer.
20. Yucel N, Sukru Erkal H, Sinem Akgun F, Serin M. Characteristics of
the admissions of cancer patients to emergency department.