Objectives
Attempts to understand patient safety using administrative data in Korea have been rare. This study develops a Korean patient safety incident code classification system and identifies its characteristics to boost diagnosis code usage for assessing patient safety.
Methods
Based on existing literature, we selected Korean Standard Classification of Diseases 7 codes for characterizing patient safety incidents using diagnosis codes. We conducted 2 rounds of review to evaluate the codes applicability to different patient safety incidents using the Delphi method. The verified diagnosis codes were then classified by incident type.
Results
Of the 54,259 Korean Standard Classification of Diseases 7 codes, 4509 were applicable for Korean patients, which were divided into 2435 code groups and 2074 candidate groups. The codes were classified into 6 categories (diagnosis, medication, patient care, operation or procedure, infection related, and other) and then further classified into 35 subcategories. The major categories of patient safety incidents, in the order of frequency, involved medication, fluid and blood related (1719, 38.1%), operation and procedure related (1339, 29.7%), and patient care related (991, 22.0%). Meanwhile, there were only 2 codes related to diagnosis.
Conclusions
Our study provides a basis for estimating patient safety incidents using diagnosis codes. We suggest that gradually increasing the utilization and accuracy of the patient safety incident codes will help develop effective patient safety indicators in Korea similar to other countries. Moreover, clinicians are also needed to be aware of using the developed code classification system.
Key Words: patient safety, patient safety incident, international classification of diseases, clinical coding, adverse event
Patient safety is the most basic requirement of providing adequate health care.1,2 However, it is not easy to obtain a comprehensive understanding of patient safety level. Thus, various methodologies such as analyses of medical malpractice litigation,3 reporting and learning systems for patient safety incidents,4 assessments of administrative data,5,6 and review of medical records7 have been used. Each method has its own strengths and weaknesses, and their interpretations vary depending on the type of patient safety incident.8 This means that there is no one-size-fits-all approach to quantifying the level of patient safety. It also means choosing an appropriate method according to the purpose of the measurement.
Among the various methods for identifying the patient safety status, the method of analyzing administrative data is an approach that can comprehensively examine the patient safety status at the population level.9 Approaches to this analysis such as using diagnosis codes for examining insurance claims are relatively inexpensive and easy to perform compared with other methods. Hence, studies examining patient safety by analyzing administrative data have been used to distinguish the various types of patient safety incidents, such as drug adverse effects and healthcare-associated infections.10,11
Furthermore, monitoring patient safety incidents using diagnosis codes can also serve as a tool for comparing patient safety in different countries. Attempts to quantify the burden of disease caused by patient safety incidents have been based on methods estimating the burden for different modeling methods.12,13 In these methods, diagnosis codes play a critical role in defining the incidence or prevalence of patient safety incidents.14 The aim to compare the global disease burden caused by patient safety incidents implies a need for a monitoring system using diagnosis codes.
However, it is well known that international discrepancies in interpreting and using the International Classification of Diseases (ICD).15,16 Previous studies to identify the full range of patient safety incidents using administrative data are limited, and they used their own diagnosis codes in each country such as ICD, Ninth Revision, Clinical Modification (ICD-9-CM),17ICD Tenth Revision, Australian Modification (ICD-10-AM),18 and ICD Canadian Version of 10th Revision (ICD-10-CA).19 Therefore, it is necessary to develop a patient safety incident code classification system in each country and continue to research and develop methods for better comparison by mapping the codes.
In Republic of Korea (hereinafter Korea), there have been attempts to understand patient safety using administrative data. Some studies have aimed at examining the occurrence of 6 types of patient safety incidents using the ICD-10 codes found in the National Health Insurance Service–National Sample Cohort5 and introducing patient safety indicators using the Korean National Hospital Discharge In-depth Injury Survey.20 However, there is still need for conducting a comprehensive exam of patient safety using administrative data. In particular, it is necessary to review whether the administrative diagnosis codes used in previous studies sufficiently cover the various types of patient safety incidents.
Thus, this study aimed to develop and identify the characteristics of a Korean patient safety incident code classification system to more easily identify whether a patient safety incident has occurred using diagnosis codes.
METHODS
The Concept of the Study
In this study, according to the definition stipulated by the World Health Organization, a patient safety incident was defined as “an event or circumstance that could have resulted, or did result, in unnecessary harm to a patient.”21 That is, the concept of a patient safety incident includes both adverse events and near misses.
This study also assumed that there were data for the present on admission (POA) that identified whether the diagnosis code existed before the time of admission,22 and if the diagnosis code is considered to have occurred after admission, the code was classified as a patient safety incident. In these cases, the POA of the diagnosis code is tagged as “N.”
Figure 1 shows the logical structure that allows us to determine whether a patient safety incident has occurred using the Korean patient safety incident code classification system derived from this study.
FIGURE 1.
Process of determining whether a patient safety incident has occurred based on the patient safety incident code classification system.
Search and Review of Candidate Patient Safety Incident Codes
One researcher mapped the ICD-10 codes established in previous studies to identify patient safety incidents from 54,259 Korean Standard Classification of Diseases 7 (KCD-7) codes and selected 4557 codes for analysis after eliminating the duplicates.10,14,19,23 The selected codes were evaluated for the probability score as a patient safety incident as follows: (1) patient safety incident highly likely, (2) patient safety incident likely, (3) possible patient safety incident, (4) previously reported as patient safety incident (cannot exclude the possibility of patient safety incident), and (5) no KCD-7 code available. The codes included after review were used in the Delphi review described hereinafter as candidate codes for the Korean patient safety incident code classification system.
Delphi Review of Candidate Patient Safety Incident Codes
The Delphi review of the candidate codes was conducted by experts with extensive research and practical experience in patient safety and the KCD-7 code classification system. A total of 5 experts participated in the review: 1 emergency medicine specialist, 1 preventative medicine specialist, 2 nursing professors, and 1 health information manager.
Before beginning the review, a meeting was held with the experts to introduce the study content. The first round of the Delphi review involved an anonymized evaluation of the probability of a patient safety incident of the Korean patient safety incident candidate codes. The reviewers used a 4-point rating scale to respond to the following questions: How appropriate is the probability score assigned to the code as a patient safety incident? How much do you agree with that score? The rating scale consisted of the following: strongly agree (1 point), agree (2 points), disagree (3 points), and strongly disagree (4 points). If applicable, reviewer comments for revisions on the probability score as a patient safety incident were collected separately.
In the second round of review, the results from the first round were anonymously shared with the reviewers, giving them the opportunity for suggesting revisions. Each round of the review lasted for approximately 1 month, and on average, 2 reminder emails were sent to the reviewer per round. Reviewer agreement was operationally defined as cases in which the coefficient of variation of the response was less than 0.5. One researcher reviewed and organized the results of the 2 rounds of the Delphi review and developed the Korean patient safety incident codes.
Categorization of Patient Safety Incident Codes
The researchers classified the Korean patient safety incident codes identified through the Delphi review based on the categories used in previous studies.7,18,19,24–27 The codes were classified into the most appropriate category based on the categorical name, description, and reviewer comments. For instance, if the categorical name, description, or comments included the key word, “drug-induced,” the code was classified as a drug-related complication.
In addition, to increase the use of this classification system, the Korean patient safety incident codes were divided into code groups and candidate groups based on the probability score as a patient safety incident. The codes with probability score of 3 or lower were included as the code groups of the Korean patient safety incident code, while those with a score of 4 were included as the candidate groups. Forty-eight codes with the probability score of 5 were excluded, because they could be found in the ICD-10 codes but not in the KCD-7.
Ethics Committee Approval
This study was approved by the institutional review boards of each participating hospital and of the National Evidence-Based Healthcare Collaborating Agency (approval number: NECAIRB18-020-10).
RESULTS
The Korean Patient Safety Incident Code Classification System
The Korean patient safety incident codes were classified into 6 categories and 35 subcategories (Table 1). The entire list of codes can be found in the supplement (Supplemental File 1, http://links.lww.com/JPS/A529).
TABLE 1.
The Korean Patient Safety Incident Code Classification System
| Category | Subcategory |
|---|---|
| 1. Diagnosis-related incidents | 1.1 Non-administration of surgical and medical care |
| 2. Medication-, fluid-, and blood-related incidents | 2.1 Infusion/injection complications |
| 2.2 Drug-related complications | |
| 2.3 Transfusion-related complications | |
| 2.4 Others | |
| 3. Patient care–related incidents | 3.1 Decubitus ulcer |
| 3.2 Delirium | |
| 3.3 Falls | |
| 3.4 Patient accidents | |
| 3.5 Sequelae of events | |
| 3.6 Others | |
| 4. Operation- and procedure-related incidents | 4.1 Anesthesia-related complications |
| 4.2 Hemorrhage-related complications | |
| 4.3 Puncture/amputations/injuries during the procedure | |
| 4.4 Foreign body or substance left following the procedure | |
| 4.5 Disruption and infection of operation wound | |
| 4.6 Pregnancy- and childbirth-related complications | |
| 4.7 Implants-related complications | |
| 4.8 Procedure-related complications | |
| 4.9 Post-procedural complications | |
| 4.10 Medical device–related complications | |
| 4.11 Radiation-related complications | |
| 4.12 Other complications | |
| 5. Infection-related incidents | 5.1 Sepsis |
| 5.2 Pneumonia | |
| 5.3 Intestinal infectious diseases | |
| 5.4 Urinary tract infection | |
| 5.5 Infectious agents | |
| 5.6 Other infections | |
| 6. Others | 6.1 Endocrine, nutritional, and metabolic disease |
| 6.2 Circulatory system diseases | |
| 6.3 Respiratory system diseases | |
| 6.4 General symptoms and signs | |
| 6.5 Patient status |
The Distribution of Korean Patient Safety Incident Codes Into Categories
There are 4509 Korean patient safety incident codes, of which 2435 belong to code groups (54.0%) and 2074 to candidate groups (46.0%; Table 2). The major categories of patient safety incidents, in the order of frequency, involved medication, fluid- and blood-related (1719 codes, 38.1%), operation- and procedure-related (1339 codes, 29.7%), and patient care–related (991 codes, 22.0%). Meanwhile, there were only 2 codes related to diagnosis. Regarding only the code groups, medication-, fluid-, and blood-related incidents comprised more than half of the 2435 codes (1291 codes, 53.0%); 973 codes (40.0%) were medication-, fluid-, and blood-related incidents, and 53 (2.2%) codes were related to patient care.
TABLE 2.
Patient Safety Incident Categories With Code Numbers of PSI Codes
| Categories | Total | Code Group | Candidate Group | |||
|---|---|---|---|---|---|---|
| n | % | n | % | n | % | |
| 1. Diagnosis-related incidents | 2 | 0.0 | 2 | 0.1 | 0 | 0.0 |
| 2. Medication-, fluid-, and blood-related incidents | 1719 | 38.1 | 1291 | 53.0 | 428 | 20.6 |
| 3. Patient care-related incidents | 991 | 22.0 | 53 | 2.2 | 938 | 45.2 |
| 4. Operation- and procedure-related incidents | 1339 | 29.7 | 973 | 40.0 | 366 | 17.6 |
| 5. Infection-related incidents | 203 | 4.5 | 108 | 4.4 | 95 | 4.6 |
| 6. Others | 255 | 5.7 | 8 | 0.3 | 247 | 11.9 |
| Total | 4509 | 100.0 | 2435 | 100.0 | 2074 | 100.0 |
PSI, patient safety incident.
Distribution of Korean Patient Safety Incident Codes Into Subcategories
Table 3 presents the Korean patient safety incident codes as subcategories. In the case of diagnosis-related patient safety incidents, there was only 1 subcategory—“nonadministration of surgical and medical care”—and 2 codes, both of which were in the code group. In medication-, fluid-, and blood-related incidents, “drugs-related complications” was the most common category, having 1119 codes. In contrast, “infusion/injection complications” consisted of only 35 codes, but all were included in the code group. Patient care–related incidents were mainly classified as patient accidents, but only 29 codes were included in the code group. Operation- and procedure-related incidents involved 333 “postprocedural complications” and 225 “pregnancy and childbirth-related complications.” Not only are these numbers of codes quite substantial, but most of the codes in these subcategories were classified into the code group, which indicated a high probability of patient safety incidents. Although there were 202 codes in “puncture/amputations/injuries during procedures,” only around a quarter of these were classified into the code group. For infection-related incidents, “other infections” were the most common with 119 codes. Although 26 codes were identified for “sepsis,” a hospital-acquired infection, all of them were classified as candidates. Meanwhile, all 21 codes for “pneumonia” were classified as codes. For “other” patient safety incidents, many of the codes were pertaining to diseases or patient conditions, and most were classified as candidate codes.
TABLE 3.
Patient Safety Incident Subcategories With Code Numbers of PSI Codes
| Subcategories | Total | Code Group | Candidate Group | |||
|---|---|---|---|---|---|---|
| n | % | n | % | n | % | |
| Diagnosis-related incidents | ||||||
| Nonadministration of surgical and medical care | 2 | 0.0 | 2 | 0.1 | 0 | 0.0 |
| Medication-, fluid-, and blood-related incidents | ||||||
| Infusion/injection complications | 35 | 0.8 | 35 | 1.4 | 0 | 0.0 |
| Drug-related complications | 1119 | 24.8 | 950 | 39.0 | 169 | 8.1 |
| Transfusion-related complications | 15 | 0.3 | 14 | 0.6 | 1 | 0.0 |
| Others | 550 | 12.2 | 292 | 12.0 | 258 | 12.4 |
| Patient care–related incidents | ||||||
| Decubitus ulcer | 14 | 0.3 | 14 | 0.6 | 0 | 0.0 |
| Delirium | 14 | 0.3 | 5 | 0.2 | 9 | 0.4 |
| Falls | 3 | 0.1 | 0.0 | 3 | 0.1 | |
| Patient accidents | 955 | 21.2 | 29 | 1.2 | 926 | 44.6 |
| Sequelae of events | 1 | 0.0 | 1 | 0.0 | 0 | 0.0 |
| Others | 4 | 0.1 | 4 | 0.2 | 0 | 0.0 |
| Operation- and procedure-related incidents | ||||||
| Anesthesia-related complications | 74 | 1.6 | 73 | 3.0 | 1 | 0.0 |
| Hemorrhage-related complications | 11 | 0.2 | 3 | 0.1 | 8 | 0.4 |
| Puncture/amputations/injuries during procedure | 202 | 4.5 | 36 | 1.5 | 166 | 8.0 |
| Foreign body or substance left following the procedure | 21 | 0.5 | 20 | 0.8 | 1 | 0.0 |
| Disruption and infection of operation wound | 5 | 0.1 | 3 | 0.1 | 2 | 0.1 |
| Pregnancy- and childbirth-related complications | 225 | 5.0 | 221 | 9.1 | 4 | 0.2 |
| Implant-related complications | 135 | 3.0 | 97 | 4.0 | 38 | 1.8 |
| Procedure-related complications | 63 | 1.4 | 33 | 1.4 | 30 | 1.4 |
| Post-procedural complications | 331 | 7.3 | 320 | 13.1 | 11 | 0.5 |
| Medical device–related complications | 67 | 1.5 | 67 | 2.8 | 0 | 0.0 |
| Radiation-related complications | 24 | 0.5 | 23 | 0.9 | 1 | 0.0 |
| Other complications | 181 | 4.0 | 77 | 3.2 | 104 | 5.0 |
| Infection-related incidents | ||||||
| Sepsis | 26 | 0.6 | 0.0 | 26 | 1.3 | |
| Pneumonia | 21 | 0.5 | 21 | 0.9 | 0 | 0.0 |
| Intestinal infectious diseases | 16 | 0.4 | 3 | 0.1 | 13 | 0.6 |
| Urinary tract infection | 2 | 0.0 | 2 | 0.1 | 0 | 0.0 |
| Infectious agents | 19 | 0.4 | 9 | 0.4 | 10 | 0.5 |
| Other infections | 119 | 2.6 | 73 | 3.0 | 46 | 2.2 |
| Others | ||||||
| Endocrine, nutritional, and metabolic disease | 79 | 1.8 | 6 | 0.2 | 73 | 3.5 |
| Circulatory system diseases | 73 | 1.6 | 2 | 0.1 | 71 | 3.4 |
| Respiratory system diseases | 4 | 0.1 | 0.0 | 4 | 0.2 | |
| General symptoms and signs | 4 | 0.1 | 0.0 | 4 | 0.2 | |
| Patient status | 95 | 2.1 | 0.0 | 95 | 4.6 | |
Distribution of Korean Patient Safety Incident Codes Through KCD-7 Classification
The code and candidate groups comprised 4.5% (2435 codes) and 3.9% (2074 codes) of the 54,259 KCD-7 codes, respectively (Table 4). More specifically, an exam of the distribution of the Korean patient safety incident codes through KCD-7 classification revealed that “injury, poisoning, and other specific outcomes (S00–T98)” was the most common at 27.7%, followed by “extrinsic cause of morbidity and mortality (V01–Y98)” (16.1%) and “diseases of the musculoskeletal system and connective tissue (M00–M99)” (11.2%). An exam of the Korean patient safety incident codes through the KCD-7 classification demonstrated that 14.0% of the “pregnancy- and childbirth-related complications (O00–O99)” were considered codes, which made up the highest proportion. The second most prevalent KCD-7 classification was “specific conditions originating from the perinatal period (P00–P96).” With reference to the KCD-7 classification, 14.8% of the candidate codes in “factors affecting health status and exposure to health services (Z00–Z99)” and 14% of the candidate codes in “injury, poisoning, and other specific outcomes (S00–T98),” which made up a high proportion of the candidate group.
TABLE 4.
Korean Standard Classification of Diseases 7 Categories With Code Numbers Included in PSI Codes
| Category | KCD-7 Codes | PSI Codes | ||||||
|---|---|---|---|---|---|---|---|---|
| Total | Code Group | Candidate Group | ||||||
| n | n | % Category Included | n | % Category Included | n | % Category Included | ||
| 1 | Infectious and parasitic diseases (A00–B99) | 2212 | 140 | 6.3 | 88 | 4.0 | 52 | 2.4 |
| 2 | Neoplasms (C00–D48) | 2084 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 |
| 3 | Blood diseases (D50–D89) | 572 | 73 | 12.8 | 62 | 10.8 | 11 | 1.9 |
| 4 | Endocrine (E00–E90) | 3054 | 194 | 6.4 | 121 | 4.0 | 73 | 2.4 |
| 5 | Mental and behavioral (F00–F99) | 1669 | 183 | 11.0 | 42 | 2.5 | 141 | 8.4 |
| 6 | Nervous system (G00–G99) | 1049 | 91 | 8.7 | 83 | 7.9 | 8 | 0.8 |
| 7 | Eye and adnexa (H00–H59) | 1104 | 105 | 9.5 | 25 | 2.3 | 80 | 7.2 |
| 8 | Ear and mastoid process (H60–H95) | 397 | 13 | 3.3 | 12 | 3.0 | 1 | 0.3 |
| 9 | Circulatory system (I00–I99) | 1235 | 167 | 13.5 | 75 | 6.1 | 92 | 7.4 |
| 10 | Respiratory system (J00–J99) | 1002 | 88 | 8.8 | 56 | 5.6 | 32 | 3.2 |
| 11 | Digestive system (K00–K93) | 2162 | 248 | 11.5 | 219 | 10.1 | 29 | 1.3 |
| 12 | Skin and subcutaneous tissue (L00–L99) | 887 | 76 | 8.6 | 58 | 6.5 | 18 | 2.0 |
| 13 | Musculoskeletal system (M00–M99) | 17,455 | 505 | 2.9 | 417 | 2.4 | 88 | 0.5 |
| 14 | Genitourinary system (N00–N99) | 1279 | 86 | 6.7 | 54 | 4.2 | 32 | 2.5 |
| 15 | Pregnancy and childbirth (O00–O99) | 1392 | 199 | 14.3 | 195 | 14.0 | 4 | 0.3 |
| 16 | Perinatal (P00–P96) | 724 | 99 | 13.7 | 90 | 12.4 | 9 | 1.2 |
| 17 | Congenital abnormalities (Q00–Q99) | 1634 | 3 | 0.2 | 3 | 0.2 | 0 | 0.0 |
| 18 | Symptoms NEC (R00–R99) | 1341 | 61 | 4.5 | 13 | 1.0 | 48 | 3.6 |
| 19 | Injuries (S00–T98) | 6067 | 1247 | 20.6 | 399 | 6.6 | 848 | 14.0 |
| 20 | External causes of morbidity and mortality (V01–Y98) | 5238 | 727 | 13.9 | 420 | 8.0 | 307 | 5.9 |
| 21 | Factors influencing health status (Z00–Z99) | 1355 | 204 | 15.1 | 3 | 0.2 | 201 | 14.8 |
| 22 | Codes for special purpose (U00–U99) | 347 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 |
| Total | 54,259 | 4509 | 8.3 | 2435 | 4.5 | 2074 | 3.8 | |
PSI, patient safety incident.
DISCUSSION
This study developed a Korean patient safety incident code classification system and examined its characteristics. Among 54,259 KCD-7 codes, 4509 were identified as Korean patient safety incident codes after a literature review, Delphi review, and deliberation among the researchers. The identified codes were then further classified into a code group (2435 codes, likely patient safety incident codes) and a candidate group (2074 codes, used as patient safety incident codes in other studies). Until now, there has not been any classification system for diagnosis codes in Korea that can comprehensively monitor the status of patient safety incidents. Indeed, this study is significant, in that the study outcomes can serve as a key clue for judging patient safety incidents using diagnosis codes.
Despite the strengths of diagnosis code-based methods in examining the incidence of patient safety incidents,8,9 there has been a paucity of studies that review and classify diagnosis codes to obtain a comprehensive understanding of the occurrence of patient safety incidents. This study developed a Korean patient safety incident code classification system through a rigorous review process. First, a literature review was conducted to identify the various adverse drug events, including a systematic review of the ICD-10 codes,10 and the studies that aimed to obtain a comprehensive understanding of patient safety incidents using diagnosis codes.14,18,19,23,26 Subsequently, the literature was reviewed by several experts from various fields experienced in patient safety and the Korean patient safety incident code classification system using the Delphi method. This methodology may be applied in future studies for developing and revising the patient safety incident code classification systems for other countries.
If any of the diagnosis codes belong to the code group of the patient safety incident code classification system, it may be possible to question whether a patient safety incident has occurred (Fig. 1). However, if the POA of the code is “N,” it will be a patient safety incident that occurred during this hospitalization period. For example, even if a patient with decubitus ulcer was assigned the code “L890,” if the POA for the code was “Y,” it would be considered a previously occurring and not yet treated decubitus ulcer. If the POA is “N,” additional characteristics of the patient safety incident should be identified through other additional methods, such as medical record review. In some cases, the degree of harm to the patient or the presence or absence of an error can be inferred from the title of diagnosis code, but in most cases, it is difficult to grasp the specific characteristics of the patient safety incident. This can be seen as a limitation of administrative data that does not contain clinical context.8,9
To monitor patient safety using diagnosis codes, it is necessary to not only develop a code classification system but also increase its utilization and ensure its accuracy.28–30 This means that clinicians who would use the patient safety incident codes must be aware of the existence of such codes and implement them in the diagnosis process in the event of an incident. Furthermore, POA codes, which contain information that enables an estimation of the time at which the diagnosis code was generated, must be inputted accurately. Nevertheless, it is presumed that the accuracy of the POA codes is rather low, in addition to the fact that physicians often overlook inputting patient safety incident codes.6 Thus, to increase the utility of the code classification system developed in this study, the accuracy of patient safety incidents and POA code recordkeeping must first be achieved.
Suggestions to improve recordkeeping may include using indicators that evaluate the accuracy of patient safety incidents or POA codes, rather than those evaluating the occurrence of patient safety incidents. In addition, it is necessary to conduct regular education on the correct input of patient safety incident codes and POA codes for healthcare professionals, especially clinicians. Notwithstanding some controversies, as in the United States, not paying compensation for certain patient safety incidents can also be a strategy to increase coding accuracy.31,32
In particular, to ensure the accuracy of patient safety incidents and POA codes, it is important to match the classification system with investigations using other patient safety monitoring methods. The classification system developed in this study is consistent with that of the Korea National Patient Safety Incidents Inquiry Survey using the medical record review.7 The accuracy of the patient safety incident and POA codes may be evaluated through codes related to patient safety incidents in the diagnosis codes of patients identified in the National Patient Safety Incidents Inquiry Survey, along with determining whether the POA code is “N.”6
The scarcity of diagnosis-related patient safety incident codes is one major limitation of the developed patient safety incident code classification system. Among the 4509 patient safety incident codes, only 2 codes related to diagnosis were grouped under “nonadministration of surgical and medical care,” which were not enough to cover all of the various diagnosis-related patient safety incidents. There is an additional need for diagnosis-related codes, such as those related to delayed or inaccurate diagnoses. Although, in this study, the diagnosis codes currently used were classified inductively, in the future, it may be necessary to develop codes specific to patient safety incidents. In addition, if an investigation is conducted to find patient safety incidents related to diagnosis, it seems more reasonable to use other methods such as interviews with patient and medical professionals rather than administrative data analysis.24,25
Another limitation is that the patient safety incident codes developed did not quantitatively examine the incidence of patient safety incidents. Although it was presumed that the utilization and accuracy of codes related to patient safety incidents were rather low in Korea,5,6 it may be important to identify which types of codes are more or less used. In the future, it may be needed to evaluate the applicability of the patient safety incident code classification system using the National Health Insurance Service or Health Insurance Review and Assessment databases.
CONCLUSIONS
We developed a Korean patient safety incident code classification system for monitoring patient safety incidents through a rigorous review process using diagnosis codes found in administrative data. In Korea, no data have been collected on patient safety incidents, which were requested by the Organization for Economic Cooperation and Development, aside from the incidence of sepsis after abdominal surgery.33 This study provided the basis for examining the incidence of various patient safety incidents using diagnosis codes. We suggest that by gradually increasing the utilization and accuracy of the patient safety incident codes, patient safety indicators comparable with other countries may be established. Furthermore, it will be necessary to increase clinicians’ awareness regarding the code classification system developed in this study.
Supplementary Material
Footnotes
The authors disclose no conflict of interest.
This research was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute, funded by the Ministry of Health and Welfare of the Republic of Korea (grant number: HI18C2339). This research was also supported by a grant from the Korean Society for Preventive Medicine in 2021.
Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.journalpatientsafety.com).
Contributor Information
Eun Young Choi, Email: 11351@naver.com.
Jeehee Pyo, Email: eesther0517@naver.com.
Young-Kwon Park, Email: 0734714@uuh.ulsan.kr.
Sukyeong Kim, Email: sukyeong.kim@neca.re.kr.
REFERENCES
- 1.Institute of Medicine, Committee on Quality of Health Care in America . Crossing the Quality Chasm: A New Health System for the 21st Century. 1st ed. Washington, DC: National Academies Press; 2001. [Google Scholar]
- 2.World Health Organization . Patient safety: a global health priority [Internet]. Available at: https://www.who.int/patientsafety/policies/global-health-priority/en/. Accessed April 12, 2022.
- 3.Choi S Shin S Lee W, et al. Medicolegal lessons learned from thyroidectomy-related lawsuits: an analysis of judicial precedents in South Korea from 1998 to 2019. Gland Surg. 2020;9:1286–1297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shin S, Won M. Trend analysis of patient safety incidents and their associated factors in Korea using National Patient Safety Report Data (2017~2019). Int J Environ Res Public Health. 2021;18:8482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ock M Kim HJ Jeon B, et al. Identifying adverse events using international classification of diseases, tenth revision Y codes in Korea: a cross-sectional study. J Prev Med Public Health. 2018;51:15–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kim J Choi EY Lee W, et al. Feasibility of capturing adverse events from insurance claims data using international classification of diseases, tenth revision, codes coupled to present on admission indicators. J Patient Saf. 2021;18:404–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kim MJ Seo HJ Koo HM, et al. The Korea National Patient Safety Incidents Inquiry Survey: feasibility of medical record review for detecting adverse events in regional public hospitals. J Patient Saf. 2022;18:389–395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Murff HJ Patel VL Hripcsak G, et al. Detecting adverse events for patient safety research: a review of current methodologies. J Biomed Inform. 2003;36:131–143. [DOI] [PubMed] [Google Scholar]
- 9.Zhan C, Miller MR. Administrative data based patient safety research: a critical review. Qual Saf Health Care. 2003;12(Suppl 2):ii58–ii63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hohl CM Karpov A Reddekopp L, et al. ICD-10 codes used to identify adverse drug events in administrative data: a systematic review. J Am Med Inform Assoc. 2014;21:547–557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.van Mourik MS van Duijn PJ Moons KG, et al. Accuracy of administrative data for surveillance of healthcare-associated infections: a systematic review. BMJ Open. 2015;5:e008424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jha AK Larizgoitia I Audera-Lopez C, et al. The global burden of unsafe medical care: analytic modelling of observational studies. BMJ Qual Saf. 2013;22:809–815. [DOI] [PubMed] [Google Scholar]
- 13.Flott K, Fontana G, Darzi A. The Global State of Patient Safety [Internet]. London: Imperial College London; 2019. Available at: https://www.imperial.ac.uk/media/imperial-college/institute-of-global-health-innovation/GlobalStateofPS_DIGITAL_16Sep19%5B2%5D.pdf. Accessed April 12, 2022. [Google Scholar]
- 14.GBD 2017 Disease and Injury Incidence and Prevalence Collaborators . Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392:1789–1858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Otero Varela L Doktorchik C Wiebe N, et al. Exploring the differences in ICD and hospital morbidity data collection features across countries: an international survey. BMC Health Serv Res. 2021;21:308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Quan H Moskal L Forster AJ, et al. International variation in the definition of ‘main condition’ in ICD-coded health data. Int J Qual Health Care. 2014;26:511–515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hougland P Xu W Pickard S, et al. Performance of International Classification of Diseases, 9th Revision, Clinical Modification codes as an adverse drug event surveillance system. Med Care. 2006;44:629–636. [DOI] [PubMed] [Google Scholar]
- 18.Jackson TJ Michel JL Roberts RF, et al. A classification of hospital-acquired diagnoses for use with routine hospital data. Med J Aust. 2009;191:544–548. [DOI] [PubMed] [Google Scholar]
- 19.Southern DA Burnand B Droesler SE, et al. Deriving ICD-10 codes for patient safety indicators for large-scale surveillance using administrative hospital data. Med Care. 2017;55:252–260. [DOI] [PubMed] [Google Scholar]
- 20.Kim YM. Application of patient safety indicators using Korean National Hospital Discharge In-depth Injury Survey. JKAIS. 2013;14:2293–2303. [Google Scholar]
- 21.World Health Organization . Final technical report for the conceptual framework for the international classification for patient safety. Available at: https://apps.who.int/iris/handle/10665/70882 Accessed on July 22, 2022.
- 22.Glance LG Osler TM Mukamel DB, et al. Impact of the present-on-admission indicator on hospital quality measurement: experience with the Agency for Healthcare Research and Quality (AHRQ) inpatient quality indicators. Med Care. 2008;46:112–119. [DOI] [PubMed] [Google Scholar]
- 23.Yoon J Oh IH Seo H, et al. Disability-adjusted life years for 313 diseases and injuries: the 2012 Korean Burden of Disease Study. J Korean med Sci. 2016;31 Suppl 2(suppl 2):S146–S157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ock M Jo MW Choi EY, et al. Patient safety incidents reported by the general public in Korea: a cross-sectional study. J Patient Saf. 2020;16:e90–e96. [DOI] [PubMed] [Google Scholar]
- 25.Pyo J Choi EY Lee W, et al. Physicians’ difficulties due to patient safety incidents in Korea: a cross-sectional study. J Korean Med Sci. 2020;35:e118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jackson TJ Michel JL Roberts R, et al. Development of a validation algorithm for ‘present on admission’ flagging. BMC Med Inform Decis Mak. 2009;9:48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Canadian Institute for Health Information . Hospital Harm Indicator: General Methodology Notes, May 2019 [Internet]. Ottawa, ON: CIHI; 2019. Available at: https://indicatorlibrary.cihi.ca/download/attachments/10453027/Hospital%20Harm%20Indicator%20General%20Methodology%20Notes.pdf. Accessed April 12, 2022. [Google Scholar]
- 28.Kaafarani HM Borzecki AM Itani KM, et al. Validity of selected Patient Safety Indicators: opportunities and concerns. J Am Coll Surg. 2011;212:924–934. [DOI] [PubMed] [Google Scholar]
- 29.Rosen AK Itani KM Cevasco M, et al. Validating the patient safety indicators in the Veterans Health Administration: do they accurately identify true safety events? Med Care. 2012;50:74–85. [DOI] [PubMed] [Google Scholar]
- 30.Borzecki AM, Rosen AK. Is there a ‘best measure’ of patient safety? BMJ Qual Saf. 2020;29:185–188. [DOI] [PubMed] [Google Scholar]
- 31.US Centers for Medicare and Medicaid Services . Hospital compare. Available at: http://www.medicare.gov/hospitalcompare/?AspxAutoDetectCookieSupport=1. Accessed July 22, 2022.
- 32.Hospital Inpatient Quality Reporting Program. Available at: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQualityInits/HospitalRHQDAPU. Accessed July 22, 2022.
- 33.Organization for Economic Co-operation and Development . OECD health care quality review: Korea, assessment and recommendations. 2012. Available at: https://www.oecd.org/korea/49818570.pdf. Accessed April 12, 2022.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.