Clinical Studies using Diagnosis Procedure Combination (DPC) Database Hideo Yasunaga, MD, PhD Department of Health Management and Policy, Graduate School of Medicine, University of Tokyo 1
What is DPC? Japanese version of Diagnosis Related Groups Diagnosis Procedure Combination (DPC) is a diagnosis-dominant case-mix system, comprised of 18 major diagnostic categories (MDC), 520 diagnostic groups, and 2,347 case-mix groups.
18 Major Diagnosis Group MDC 01 Central nervous system MDC 02 Eyes MDC 03 Ears, Nose and Throat MDC 04 Respiratory system MDC 05 Cardiovascular system
520 Diagnosis Groups Major Diagnosis Group 05 (Cardiovascular system) 050030 Acute Myocardiac Infarction 050050 Angina Pectoris 050060 Cardiomyopathy 050080 Valvular Diseases
2,347 DPC 050030(Acute myocardiac infarction) AMI, tpa, Coronary stent 050030xx03x3xx AMI, IABP, CABG 050030xx02x4xx
The DPC Database The DPC Database is a patient discharge and administrative claims database. Data are compiled between July and December each year by the DPC Research Group (Chief: Shinya Matsuda MD, PhD) funded by the Ministry of Health, Labour and Welfare, Japan.
Administrative Claims Database in the US (1)Nationwide Inpatient Sample (NIS) Database -includes 8 million patients per year from 1,044 hospitals in 40 states, representing approx. 20% of all inpatients in the US. (2) Medicare Claim Database -includes all patients 65 yrs -linked with Surveillance, Epidemiology and End Results (SEER) program of cancer registries (SEER-Medicare Linked Database).
Japanese DPC Database includes around 3 million inpatients per half a year, representing approx. 45% of all acute care inpatients in Japan.
Data The database include patient information on -Diagnoses coded with the ICD-10 codes -Procedures -Comorbidities at admission ( 入 院 時 併 存 症 ) -Complications after admission ( 入 院 後 合 併 症 ) -Drugs and Devices used -In-hospital mortality( 在 院 死 亡 ) -Length of stay( 在 院 日 数 ) -Costs 9
-Cancer stage, Chemotherapy, Radiotherapy -Coma Scale -ASA, NYHA, Hugh-Jones, Child-Pugh -Burn Index -Pregnancy -Smoking index etc.
Clinical Statistics using the DPC Database 11
Operative Mortality following Cancer Surgery 2007 (がん 手 術 後 の 在 院 死 亡 率 ) N inhospital death % Lung 17,341 173 1.00% Esophagus 3,327 152 4.57% Stomach 31,910 512 1.60% Liver 10,026 271 2.70% Pancreas 6,123 196 3.20% Colon 27,958 609 2.18% Rectum 17,927 321 1.79% Kidney 8,413 76 0.90% Bladder 2,418 49 2.03% Prostate 7,908 4 0.05% Uterus 7,501 18 0.24%
Operative Mortality following Cancer Surgery 2007 50-59yrs 60-69yrs 70-79yrs 80yrs Lung 0.5% 0.7% 1.3% 2.1% Esophagus 2.7% 3.3% 7.3% 10.0% Stomach 0.6% 1.0% 1.8% 3.9% Liver 1.6% 2.3% 3.7% 4.5% Pancreas 1.4% 2.9% 4.0% 6.6% Colon 1.0% 1.3% 2.2% 4.6% Rectum 0.6% 1.0% 2.5% 4.6% Bladder 0.0% 1.4% 3.0% 3.1% Uterus 0.3% 0.2% 0.6% 1.0%
Surgical Volume and Operative Outcomes ( 手 術 件 数 と 手 術 成 績 ) Studies have indicated that higher operative caseload volume is associated with lower surgical mortality after complex surgical procedures. However, little is known about the relationship between operative volume and postoperative complications.
Postoperative complications following Nephrectomy for Renal Cancer (n=7988) Urinary tract infection 117 (1.5%) Ileus 103 (1.3%) Surgical site infection 97 (1.2%) Cardiac events 87 (1.1%) Sepsis 63 (0.8%) Respiratory complications 58 (0.7%) Renal failure 46 (0.6%) Venous thromboembolism 44 (0.4%) Pancreatitis 21 (0.3%) Stroke 17 (0.2%) Peritonitis 14 (0.2%) Overall, 595 patients (7.4%) had at least one complication. The in-hospital rate of death was 0.84% (n=67).
Age and Outcomes
Hospital volume and outcomes *Hospital volume: the annual number of nephrectomy perfomed at each hospital
Multivariate regression
Incidence of Malignant Hyperthermia( 悪 性 高 熱 ) Malignant hyperthermia (MH) is a life-threatening pharmacogenetic disease that occurs during general anesthesia. A potential risk of MH is the use of volatile anesthetic agents ( 揮 発 性 麻 酔 薬 ) and muscle relaxants ( 筋 弛 緩 薬 ). The true incidence of MH remains unclear because of a paucity of universal reporting.
Incidence of MH determined with the DPC database 17 MH detected among 1,238,171 patients undergoing general anesthesia. N % MH Incidence (per million) Total 1,238,171 100.0% 17 13.7 Volatile agents Sevoflurane 932,771 75.3% 14 15.0 Isoflurane 33,231 2.7% 0 0.0 Halothane 682 0.1% 0 0.0 Enflurane 35 0.0% 0 0.0 Muscle Relaxants Suxamethonium 19,871 1.6% 0 0.0 Vecuronium 782,899 63.2% 10 12.8 Pancuroneum 11,286 0.9% 0 0.0 Rocuronium 246,572 19.9% 6 24.3 Propofol 949,694 76.7% 12 12.6
DES was introduced in September 2004 in Japan. Impact of Drug-eluting Stents (DES) on Treatment Option Mix for Coronary Artery Disease in Japan Coronary artery bypass grafting (CABG) -on-pump CABG -off-pump CABG Percutaneous coronary intervention (PCI) -Bare metal stent (BMS) -Drug-eluting stent (DES)
4-year trend in proportions of therapeutic options for Angina Pectoris from 144 hospital in July, 2004-2007 PCI:percutaneous coronary intervention, CABG:coronary artery bypass grafting DES:drug eluting stents,bms:bare metal stents
DES not only replaced BMS, but also expanded patients eligibility for PCI. Increase in PCI was concurrent with a temporary reduction in on-pump CABG but in the increase in off-pump CABG. 23
The Effect of Mechanical Thromboprophylaxis for Pulmonary Embolism following Abdominal Cancer Surgery ( 腹 部 がん 手 術 後 の 肺 塞 栓 症 に 対 する 間 欠 的 空 気 圧 迫 装 置 による 予 防 効 果 ) 24
Incidence of Perioperative Pulmonary Embolism in the groups with and without mechanical prophylaxis No prophylaxis Mechanical prophylaxis N PE % N PE % Overall 18,630 38 0.20% 52,231 85 0.16% Gastrectomy 5319 5 0.09% 16070 24 0.15% Colectomy 4849 7 0.14% 15233 20 0.13% Rectal surgery 3070 9 0.29% 9003 12 0.13% Hepatectomy 1751 4 0.23% 3391 5 0.15% Nephrectomy 1474 7 0.47% 3531 13 0.37% Prostatectomy 1376 4 0.29% 3070 5 0.16% Hysterectomy 1202 3 0.25% 2917 7 0.24%
Duration of anesthesia 180 Logistic regression odds ratio 95% confidence interval Reference 181-240 1.07 0.42-2.72 0.893 241-300 2.00 0.84-4.75 0.117 301-360 2.24 0.90-5.54 0.081 >360 4.43 1.90-10.32 0.001 Thromboprophylaxis None Mechanical thromboprophylaxis Reference 0.53 0.31-0.92 0.025 p
Measles in Japan, 2007-2008 America was declared free from endemic measles transmission in 2002. In Europe, 29,000 measles infections were reported in 2004. In Japan, measles outbreaks occurred in 2001 and 2007. Measles sometimes accompanies diverse complications, including pneumonia, otitis media, and central nervous system involvement. We searched for measles-related complications and hospitalizations in Japan in 2007-2008, using the DPC Database.
Measles Complications Age (years) 14 15-29 30 All Any complication (%) 6.6 4.2 3.7 4.8 Pneumonia (%) 5.2 1.8 1.4 2.8 Encephalitis (%) 0.3 1.0 0.7 0.7 Meningitis (%) 0.0 0.1 0.5 0.2 Intestinal complications (%) 0.2 0.3 0.1 0.2 Conjunctivitis (%) 0.3 0.1 0.3 0.2 Otitis media (%) 0.2 0.1 0.0 0.1 Other complications (%) 0.4 0.8 0.7 0.6
Age Distribution Age distribution revealed two peaks in 4 and 15-29 years old. Measles-mumps-rubella (MMR)vaccine failure in late 1980s and early 1990s could be associated with the high incidence in patients aged 15-29 years.
Venomous snake bites in Japan About 15% of 3000 species of snakes worldwide are venomous, including rattlesnakes, cottonmouths and copperheads in North America, cobras, mambas and kraits in tropical zones, and mamushi (A. blomhoffii) and habu (T. flavoviridis) in Japan. Crotalidae species (pit vipers) with a triangular head, elliptical pupils, and retractable fangs (NEJM2002;347:347-56)
Mamushi and Habu bites (N=1,670) 352 2007 2008 222 223 227 171 188 131 103 23 19 7 4 July Aug Sep Oct Nov Dec
n % Region Northern area Hokkaido 8 0.5 Tohoku 71 4.3 Central area Kanto 143 8.6 Chubu 285 17.1 Western area Kinki 210 12.6 Chugoku 243 14.6 Shikoku 105 6.3 Southern area Kyushu 522 31.3 Okinawa islands 60 3.6
Snake venom increases capillary membrane permeability, which leads to extravasation of electrolytes, albumin, and red blood cells into the bite site. Edema, hypoalbuminemia, and hemoconcentration can result in hypovolemic shock and acute renal failure. n % Local complications Wound infection 35 2.1 Compartment syndrome 31 1.9 Systemic complications Shock 77 4.6 Disseminated intravascular coagulation 29 1.7 Rhabdomyolysis 52 3.1 Acute renal failure 55 3.3 Death 3 0.2
Health service researches using the DPC database Researchers can utilize the DPC database to identify, track, and analyze national trends in healthcare utilization, access, outcomes and costs. The data can be utilized to make a better allocation of healthcare resources.
Framework for clinical epidemiological studies using the DPC database Clinical experts Epidemiologists Discussion Study Design Paperwork Data extraction Data screening Analysis
Development of the DPC database Increase in the coverage rate Improvement of usability for researchers Linkage with other database -Outpatient database -Cancer registry -Long-term care database