Abstract
Patient/family complaints are a valuable source of information for providing patient-oriented healthcare. This study aimed to identify and systematize patient/family complaints about healthcare services, focusing on complaints caused by “things.” A qualitative descriptive study was designed. Open data of patient and family voices published on the website of university hospital were collected from 27 hospitals for the period June 2020 to August 2020. From the collected data, we excluded praise and compliments, and complaints regarding “people.” The results revealed 1,476 complaints, with 1,755 codes. Patient/family complaints were categorized into five domains (access to hospital or line of flow in the hospital, outpatient, inpatient, facilities/equipment, publicity/documents), 46 categories, and 150 sub-categories. A total of 545 codes were excluded to avoid duplication: [1] 253 related to hardware, [2] 222 related to operations, and [3] 70 related to maintenance. This study may provide useful data to inform future studies using patient/family complaints to improve healthcare services for hospitals aiming to provide patient-centered care.
Keywords
Introduction
Complaints made by patients and families typically express dissatisfaction as well as requests regarding health services, and have attracted attention as a valuable source of information for hospitals to improve the quality of health services, to realize “patient-oriented healthcare.”1,2 As healthcare changes from “disease-centered” to “patient-centered,” it is expected that free-text data will provide patient experiences that cannot be revealed by the closed questions used in surveys. 3 Although the incidence of patients’ complaints has recently increased, 4 these complaints should not be seen as entirely negative for hospitals because they constitute a valuable resource for organizations to improve their healthcare services. 5 The patient's perspective is the most important aspect of healthcare services. To understand patient-oriented healthcare services, systems must be built to collect, organize, analyze, and respond to patient/family complaints. 6 Taxonomy of the customer's voice is positioned as one of the key items for system building in complaint management. 6
A taxonomy of patient complaints (three domains [“clinical,” “management,” “relationship”], seven categories, and 26 subcategories) was recently developed by a systematic review. 7 However, more than 90% of the articles included in the review were conducted in Western countries, and some reports were found to lack definitions or explanations of terms.8,9 Differences in the way the terminology is perceived can lead to misunderstandings between different populations. Many East Asian countries such as Japan, Korea, and Taiwan have free-access health care systems, whereas many Western countries have home-doctor systems. According to the taxonomy of previous studies, 7 treatment is considered to be distributed in terms of domains and categories. 10 For example, it has been reported that “medication errors” are sometimes categorized as complaints caused by “things” or as “inappropriate communication” caused by “people”. In previous taxonomies,11–13 it is unclear whether the cause of the complaint is “people” or “things,” making it difficult to identify appropriate improvements. To avoid confusion in the implementation of the taxonomy, it is necessary to analyze patient and family complaints related to “people” and complaints related to “things” separately. 14 Importantly, policies and evaluation methods for improving the quality of health care services differ between “people” and “things.” Direct comparisons with previous studies in different countries can be difficult and the classification of complaints in each country needs to be considered.
Complaints caused by “people” (ie, doctors, nurses, and other staff working in the hospital) are often reported as complaints about communication, attitude, and treatment/nursing practices.8,15–17 The systematic review also indicated that the percentage of complaints caused by “people,” such as treatment, communication, staff attitudes, skill and conduct, was higher than that caused by “things,” such as the hospital environment, finance, and billing. Complaints caused by “people” are difficult to decrease over time, even when time is taken to make improvements. 14 Complaints caused by “things” are reported to be fewer in number than those caused by “people.”8,16 However, a study conducted at a Japanese institution found that complaints caused by “things” were more common than those caused by “people.” 14
The current study sought to clarify and systematize patient/family complaints about health services, focusing on complaints caused by “things” to understand the complaints from the patient/family perspective and to identify areas for improvement. By systematizing patient/family complaints, hospital staff can understand the needs of patient/family that are not clearly expressed, such as the shame and resignation that patients/families feel when they want to communicate something to hospital staff but are unable to do so, and uncertainty about who to talk to. In particular, hospital facilities and environments have evolved over time, and it is possible to gain an understanding of the points raised by patients/families, enabling the efficient consideration of measures to respond to them. Such rapid response can lead to improved patient/family satisfaction and increased quality of life. Additionally, the hospital can quickly improve healthcare services from the perspective of patients/families, thereby gaining the trust of patients and families and making it possible to create more sustainable hospitals.
Methods
Design
The study adopted a qualitative descriptive design and reported according to the integrated criteria for reporting qualitative research (COREQ) checklist. 18
Data Source
From the database of 8,298 hospitals (as of August 5, 2020) included in the list of members of the Japan Hospital Association, we selected target hospitals using the following inclusion criteria: To capture diversity with regard to complaints, it was necessary to examine a large hospital with as many departments, facilities, and equipment types as possible (with more than 500 beds), and which publicizes patient letters. There are no clear standards for publicizing patient voice in Japan. However, there is a system where a third-party organization evaluates hospital functions. One of the evaluations was the requirement to collect and provide feedback on patient/family requests. Some hospitals publicize these requests for the purpose of meeting this standard, but the criteria for such public disclosure are not clear, so it is left to each hospital.
To check whether patient letters were publicly available, we browsed each hospital's website and searched using the keywords “opinion” and “patient letter” using the hospital's internal search function of the website. We excluded specialty hospitals, because we sought to identify a wide range of patient/family perspectives in terms of age and disease.
For hospitals that met the above criteria, patient letters (secondary data not including personal information) publicized on the website were collected as data for this study. In other words, the patient letters used in this study were data collected independently by each hospital from patient/family for the purpose of improving the quality of each hospital, not collected newly for our study. Because examining patient letters is an indirect method of collection, this approach was judged to be effective for capturing the essence of complaints. Because a high level of kindness and politeness are considered to be national characteristics of public behavior in Japan, 19 we assumed that complaint letters may allow individuals to express feelings that they do not usually feel able to express. Among the publicly available patient letters, complaints regarding “things” were included in the analysis, and 703 compliments and 493 complaints regarding “people” were excluded.
Data Analysis
The analysis adopted Elo and Kyngas’ analysis method on the basis of Berelson's content analysis, 20 because the highest priority was to retain the statements expressed by the patient/family in their own words, and to reflect as closely as possible the content articulated in the text without inference. In addition to Researcher A, a medical faculty member (Researcher B) with experience conducting qualitative research with a PhD in nursing was asked to serve as the analyst. Individual sentences in each of the complaints to be analyzed were extracted as a “unit of record,” 21 which was defined as “a sentence that contains one content as one item.” To integrate the recognition of the unit of record between two researchers, it is recommended that “analysis be conducted by multiple researchers as a guarantee of objectivity” and that the “analysis be conducted while checking for reliability in the analysis procedure.” 22 Therefore, we first conducted a preliminary confirmation process by randomly selecting 20 cases from among the patient/family letters for checking whether the perception of coding showed agreement between the two researchers. After confirming that their perceptions were consistent in the analysis, all of the remaining data were independently coded by the two researchers. After both researchers completed coding, they checked with each other to determine the number of codes and the sections that had been coded for patient/family letters.
Complaints that did not show agreement regarding the number of coding and sections of coding were discussed again until they showed agreement between the two researchers. All codes were then categorized by gradually increasing the level of abstraction in the order of sub-category, category, and domain.
To confirm the reliability of the analysis between two researchers, we calculated the agreement rate in terms of the number of codes counted in a single complaint. The calculation method used the intraclass correlation coefficient, 23 and the result was 0.946, which was classified as “mostly consistent.”
Results
Of the 8,298 hospitals listed by the Japan Hospital Association, 293 hospitals had more than 500 beds, and 59 hospitals published patient/family letters on their hospital websites. Of these, 2,672 patient/family letters were published from 2006 to 2020 for 27 university hospitals, excluding specialty hospitals. The number of complaints regarding “things” was 1476 (75.0%). The average number of characters was 82.96 characters (range: 7-1,313). Over time, the number of complaints regarding “things” per hospital increased each year (Table 1). Two researchers conducted an intercept for each complaint: 1,728 for Researcher A and 1,761 for Researcher B. There were 33 letters for which the number of interceptions did not show agreement. These 33 letters were discussed among the researchers and consensus was reached on a final code count of 1,755.
Average Number of Complaints Regarding “Things” Per Hospital.
Duplication.
Complaint codes were categorized into five domains (access to hospital or line of flow in the hospital, outpatient, inpatient, facilities/equipment, publicity/documents). The five domains were 46 categories, 150 sub-categories, and 545 codes. Hereafter, raw data are indicated by quotation marks (“”), sub-categories by square brackets ([]), and categories by angle brackets (<>).
Access to hospital or line of flow in the hospital included six categories (Location, Public Transportation, Roads, Parking, Bicycle parking, Flow line in the hospital) and 15 sub-categories (Table 2). Patient/family observed safe and smooth access to healthcare services and sought convenience in accessing the hospital not only inside the hospital, but also outside the hospital. For example, in the <roads > category, the raw data “I would like to have an information board from the crossroads of A to the university hospital. I sometimes couldn’t remember which way to go” was included in the [off-site roads] sub-category. In the < public transport > category, the raw data “There is a persistent lack of taxis between 6 and 8
Taxonomy of Domain “Access to the Hospital or Flow Lines in the Hospitala” (Excerpt).
Complaints regarding transportation to the hospital and things used in line of flow in the hospital.
The code is divided into three parts (Hardware, Operation, Maintenance) for convenience.
Outpatients included seven categories (Reception, Examination, Medical examination, Consulting room/Meeting Room, Accounting, Pharmacy, Waiting Area in hospital) and 27 sub-categories (Table 3). Patient/family were observed to experience inconvenience and discomfort in the outpatient flow of reception - examination - medical care - accounting - pharmacy, with complaints about waiting times. For example, in the < medical examination > category, the raw data “Reconsider the way patients carry around the file of outpatient calculations for a long time.” was included in the [medical examination environment] sub-category. In the <accounting > category, the raw data “A bank credit card was used at the after-hours reimbursement location, but it did not work” was included in the [automatic payment] sub-category.
Taxonomy of Domain “Outpatienta” and “Inpatientb” (Excerpt).
Complaints about places and things used by patient/family in outpatient care.
Complaints about places and things used by the patient/family during hospitalization.
The code is divided into three parts (Hardware, Operation, Maintenance) for convenience.
Inpatient included eight categories (Ward, Hospital room, Hospitalization Procedures, Day room, Surgeon waiting room, Shower room / Bathroom, Laundry room, Hospital food) and 39 sub-categories (Table 3). Patients/families were observed for complaints they felt before and during hospitalization. In the < patient room > category, the raw data “I have to stand up every time I want to insert a card to watch TV” was included in the [TV/TV cards/radios] sub-category. In the same category, the raw data “To be honest, it is very time-consuming to fill in my name and other information to get my visiting badge at the entrance every day” was included in the [visiting] sub-category.
Facilities/equipment included 17 categories (Entrance, Toilet, Restaurant, Beverage vending machines, ATM, Store, Library / Information Corner, Non-Smoking / Smoking, Nursing Room, Barber, Post, Child Related Facilities, Patient recreational facilities, Shared equipment, Environment, Exterior, Services) and 56 sub-categories (Table 4). Patient/family described convenience and discomfort in terms of requests for the establishment of things and, if established, dissatisfaction with improvements and maintenance. For example, in the <toilet > category, the raw data “There is only a handrail on one side” and “Sensor does not respond well and does not flow” were included in the [dissatisfaction] sub-category. In the < patient room > category, the raw data “I have to stand up every time I want to insert a card to watch TV” was included in the [TV/TV cards/radios] sub-category. In the <store > category, the raw data “I want the store to be wider, with more products, and less varieties of bread” was included in the [assortment] sub-category.
Taxonomy of Domain “Facilities / Equipmenta” and “Publicity / Documentsb” (Excerpt).
Complaints about places and things shared by patient/family regardless of whether they are outpatients or inpatients.
Complaints about information about the hospital or documents that patient/family see or use.
The code is divided into three parts (Hardware, Operation, Maintenance) for convenience.
Publicity/documents included eight categories (Name tags, hospital evaluation, Suggestion box, Home Page, Postings / issues, Hospitalization Brochure, Patient group, Document preparation) and 13 sub-categories (Table 4). Patient/family were observed to demand detailed attention to the objects they use. For example, in < suggestion boxes > category, the raw data “How do patients’ letters and requests reach the hospital director?” and “How are the letters discussed in the hospital after we post them?” were included in the [utilization] sub-category. In the <nametags > category, the raw data “The writing on the name tag is too small to see” and “I was thinking of asking the midwife, but there are people who do not have ‘midwife’ on their nametag” were included in the [description] sub-category.
Discussion
The purpose of this study was to systematize the perspectives of patient/family complaints about healthcare services by focusing on complaints regarding “things” in patient/family letters (open ended) to university hospitals and conducting qualitative content analysis. The results revealed that patient/family using the university hospital have a variety of complaints that are difficult to surface about the things they use inside and outside the hospital, and complaints about the hardware, operation, and maintenance of the things they use were extracted.
In the current study, creating a taxonomy that reflected the patient's voice revealed where the complaints occurred, what things caused the complaints, and the conditions under which the complaints occurred. Moreover, by focusing on complaints caused by “things,” the taxonomy reflecting the patient's voice enabled the categorization of complaints by place and things rather than by theme. If the complaints were categorized by locations and things, the situation would differ from place to place in the same facility, and the causes of complaints would be more concrete. Because the collection of physical location information, such as location, is considered to lead to intra-organizational sharing and more efficient responses from the perspective of facility improvement, 5 we believe that the taxonomy of complaints by “thing” systematized by this study will help university hospitals that provide patient/family-centered healthcare.
Usefulness of Patient/Family Letters
Although questionnaire surveys using Likert scales, such as patient satisfaction surveys, provide a useful approach for quantifying customer satisfaction, the process of designing questions and surveys is time-consuming and can introduce sampling bias. 5 When a hospital uses patient/family letters, the management department must determine an investigation response policy to confirm the validity of the content, because they are one-sided written opinions and may include unreasonable demands or falsehoods. 24 Therefore, it is difficult to judge the reliability of patient and family complaints. However, the secondary data in PDF format available on the hospital's website suggested the reliability of the patient/family complaints. 25
In the current study, complaints caused by “things” accounted for approximately 75% of complaints. This finding is broadly consistent with the results of a previous study conducted at one hospital in Japan. 14 However, contrary to the previous study, 14 the number of complaints regarding “things” in the present study was found to be increasing year by year. Although we were unable to confirm how hospitals that disclose complaints respond to them, it is possible that complaints are increasing over time in Japan. Nevertheless, according to a previous study, 14 even if patient/family complaints increased, if hospitals respond in good faith, complaints about “things” will decrease compared with complaints about “people.” If each hospital can recognize that complaints constitute positive information for hospital improvement, rather than responding to complaints on an ad hoc basis, complaints can provide an effective tool for hospitals to implement appropriate responses and improvements from the patient/family perspective. Additionally, hospitals can create taxonomies of complaints, which can inform improvements in healthcare services by improving the consistency of responses and realizing preventive measures through accumulation and detailed improvements.
Taxonomy Considerations
To systematize the taxonomy of complaints, five conditions should be met, 6 which are described in relation to the current study as follows.
Regarding the “action-oriented” consideration, we devised a way to include location information in the category names and codes, because it is conceivable that the corresponding department may differ depending on the location, even within the same hospital. In this way, we can analyze the causes within the hospital and draw conclusions about measures to deal with the problem.
Regarding the “clear categorization” consideration, previous studies have not typically provided definitions of category names8,9,17,26 or have only provided abstract keywords to describe the categories.12,15,27–30 For example, the term “waiting time” is generally used in Japan to refer to the time until a patient is called for a medical examination or an accounting counter. However, in other countries with different medical systems, the term “waiting time” can refer to the number of days until a patient is seen, so the category name should be “waiting time in hospital/days. In addition, there were complaints not only about long waiting times, but also about the way people spend their waiting time, such as “I wish the staff would tell me how long I still have to wait,” “There are no magazines,” and “I wish they would provide origami for the children” during the waiting time.” Thus, “waiting time” has different causes, and is not limited to time alone. In previous studies, “access” was defined as “Any deficit in access to medical staff” 16 . In Japan, “access” is commonly interpreted as the way of getting from home to the hospital. Many hospital websites provide information about how to get to the hospital by car, train, or bus, and how long the journey takes. Therefore, we believe that it would be helpful to provide a specific description with the category name with a specific description, such as “access to hospital,” to reduce misunderstandings among researchers conducting qualitative research.
Regarding the “completeness” consideration, the taxonomy system must be able to record all patient complaints. In addition, complaints such as those regarding Internet availability (eg, free Wi-Fi) or coronavirus infection control may be influenced by the social context. The taxonomy for the complaint system should collect complaints continuously, not temporarily, and the complaints should add to the taxonomy as appropriate.
Regarding the “customer orientation” consideration, it may have been appropriate to proceed with reference to Elo and Kyngas, who described the Berelson method in more detail, which is designed to reflect what is explicitly stated in the patient/family's own voice as closely as possible.
Regarding the “convenience” consideration, a taxonomy system should be easy for anyone to understand and use. For this reason, in the current study, we tried to use one or a few words to name the categories. In doing so, we considered that this was also in accord with the need for the taxonomy to be classified in a highly compatible format so that all staff in the hospital can understand and share it equally. 5 Thus, in the current study, we created an original taxonomy of complaints, with an emphasis on the current patient/family perspective in Japan, while taking into consideration the fulfillment of the five necessary conditions described above.
Limitations
This study involved several limitations. First, the taxonomy cannot be generalized because the data were obtained from only a few university hospitals, only a few patients/families sent complaints as letters, and the complaints included environmental, customary, and cultural elements unique to Japan. It is unlikely that all hospitals publish all letters (including adverse or serious cases for the hospital that could lead to litigation), and the publicized complaints may be in relation to issues that are relatively easy to publicize. In accord with this notion, our data indicated few life-threatening safety complaints and many complaints about comfort and convenience. Expanding the scope to include medium and small hospitals and examining whether the taxonomy can be adapted may be helpful for generalizing the current findings. Second, because the data had already been processed and did not include personal information (age, gender, disease), the influences of these factors are unknown. Third, the taxonomy developed in the current study was developed and validated by two researchers and supervised by one hospital architecture expert from a patient perspective. The feasibility of using this taxonomy in other countries needs to be examined in light of the healthcare systems and social conditions in each country. Because the complaints were collected between 2006 and 2020, they contained a mixture of relatively recent complaints and those that were more than 10 years old, and older complaints may no longer be relevant in some countries. Because not all hospitals in the country are modern facilities and equipment, all complaints were included in the taxonomy in the hope that they would be of interest to all readers. In the future, consideration of complaints caused by “things” should focus on recent complaints that reflect all hospitals of interest.
Conclusions
The content of the letters written by patients/families using university hospitals revealed needs that are difficult to surface. A focus on complaints related to “things” enabled us to grasp detailed patient/family perspectives about the hardware, operation, and maintenance at the facilities. Moreover, these data are abundant and highly transferable, and can provide useful information that contributes to the improvement of the quality of healthcare services for general hospitals that aim to provide healthcare that emphasizes the patient/family perspective.
Key Points
The taxonomy of complaints in the previous study shows a mixture of complaints caused by “people” and “things.”
Complaints caused by “people” are difficult to reduce, but complaints caused by “things” can be reduced.
A taxonomy focusing on complaints caused by “things” could be developed, and complaints about hardware, operation, and maintenance aspects were identified.
It was clear that patients/families using a university hospital had complaints about various things inside and outside the facility that are difficult to surface.
Footnotes
Acknowledgements
We would like to express our sincere gratitude to Mr Hideo Kubota, former Nikken Sekkei Ltd director, first-class architect, and current representative of Kubota Design Office, a leading expert in patient-perspective hospital design, for his advice on terminology and classification hierarchy used in the creation of the taxonomy. We thank Benjamin Knight, MSc., from Edanz (
) for editing a draft of this manuscript.
Consent for Publication
Informed consent for patient information to be published in this article was not obtained because this study used only existing information (secondary data) available on the hospital's website and did not deal with information that could identify specific individuals.
Conflicts of Interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Approval
This study was approved by Osaka Dental University Ethics Committee (Approval No.111274), Japan.
Funding
This work was supported by Osaka Dental University Personal Research funding.
Japan Society for the Promotion of Science, (grant number Grant-in-Aid for Scientific Research(B) JP 22H0341).
Statement of Human and Animal Rights
No animals were used in this research. All procedures performed in studies were in accordance with the ethical standards of institutional and/or research committee and with the 1975 Declaration of Helsinki, as revised in 2013.
Standards of Reporting
COREQ guidelines were followed.