| Background:Many studies on the prognostic factors of community-acquired pneumonia in the elderly have been done, and some prediction rules have been built and widely used in clinic already to identify different risks patients which help physicians make more rational decisions in the last two decades. But these rules were all built on the dates came from westerners with modern medicine interventions. Rarely, the similar studies with combined treatment of Traditional Chinese Medicine and western medicine were reported, nor prediction rule based on clinical dates from oriental was built.Objective:To determine the prognostic factors of community-acquired pneumonia in the elderly with combined treatment of traditional Chinese medicine and western medicine, and built a new prediction rule to identify inefficacy outcome with community-acquired pneumonia in elderly.METHODS: Design: A prospective descriptive study. Setting: Four Departments of Respiratory Medicine in The First Affiliated Hospital to Changchun University of Chinese Medicine, The First Affiliated Hospital of Henan College of TCM, Jiangsu Province Hospital of TCM and The Affiliated Hospital to Shandong University of Chinese Medicine. Patients: All patients admitted in the four wards aged > or = 45 and final diagnosed as community-acquired pneumonia were enrolled in our study. Measurements: Analysis a total of 80 variables, the main outcome is the state of clinical recovery or inefficacy at the time of 14 days after admitted or discharged. All the date were collected within 24 hours after admitted, and some of them within 4 days, 7 days, and 14 days or at the end of follow-up to determine efficacy. Unscrambling the data first, includes the loss of value of alternative, multi-center differences in calibration, the handling of quantitative data (A group: no transformation; B group: classified into a binary data by the receiver operating curve (ROC) analysis; C group: classified into a multi-layered classification of data). Then all dates were divided into six different groups based on the patient's age and admission severity: that is, the middle-aged and elderly group ( > or = 45 years) in the Full Analysis Group(FAG) (400 cases of all patients), the elderly group (> or = 60 years) in the FAG, the middle-aged group (45-59 years old) in the FAG, the middle-aged and elderly group in sub-group (patients with moderate and severe CAP), the elderly group in sub-group, the middle-aged group in sub-group. Univariate analysis and multivariate analysis methods should be used to determine the prognostic factors. In the establishment of the new prognostic tool, multivariate analysis, ROC and the one-way analysis should be used.Results: Finally, a total of 400 patients had completed this study and were selected for analysis in this study. After univariate analysis and multivariate analysis, we had identified different factors associated with inefficacy in different cohorts:â…°)The main results related to patients aged > or = 45 in FAGThere was statistically significant difference in four factors: respiratory rate, anemia, confusion and different centers in series A.There was statistically significant difference in five factors: respiratory rate, hemoglobin, blood urea nitrogen, confusion and different centers in series B.There was statistically significant difference in three factors: anemia, confusion and different centers in series C.â…±)The main results related to patients aged > or = 60 in FAGThere was statistically significant difference in four factors: anemia, moist rales, different centers and PSI scores in series A.There was statistically significant difference in five factors: C reactive protein, anemia, moist rales, different centers and PSI scores in series B.There was statistically significant difference in four factors: anemia, moist rales, different centers and PSI scores in series C.â…²)The main results related to patients aged > or = 45 and < 60 in FAGThere was statistically significant difference in two factors: different centers, education degree in series A.There was statistically significant difference in three factors: neutrophil(%), blood urea nitrogen and different centers in series B. There was statistically significant difference in three factors: different centers, education degree and severity in series C.â…³)The main results related to patients aged > or = 45 in subgroupThere was statistically significant difference in four factors: systolic pressure, anemia, confusion and different centers in series A.There was statistically significant difference in six factors: body mass index, hemoglobin, blood urea nitrogen, arterial partial pressure of oxygen, confusion and different centers in series B.There was statistically significant difference in five factors: systolic pressure, arterial partial pressure of oxygen, anemia, confusion and different centers in series C.â…´)The main results related to patients aged > or = 60 in subgroupThere was statistically significant difference in six factors: respiratory rate, systolic pressure, anemia, confusion, moist rales and PSI scores in series A.There was statistically significant difference in six factors: respiratory rate, arterial partial pressure of oxygen, cost of hospitalization, anemia, confusion and moist rales in series B.There was statistically significant difference in five factors: respiratory rate, arterial partial pressure of oxygen, anemia, confusion and moist rales in series C.â…µ)The main results related to patients aged > or = 45 and < 60 in subgroupThere was statistically significant difference in albumin in series A.There was statistically significant difference in albumin in series B.There was statistically significant difference in cost of hospitalization in series C.â…¶)The progress of the new rule builtIn multivariate analysis in patients aged > or = 45 in (FAG), seven variables were associated with inefficacy: respiratory rate in series A and B, hemoglobin and blood urea nitrogen anemia in series B, anemia, confusion and different centers. Excluding the hemoglobin due to high correlation with anemia and different centers, in the further analysis with binary logistic regression, respiratory rate in series A, blood urea nitrogen in series B, anemia, confusion remained as independent variables associated with inefficacy, and consist the new prediction rule named ACUR(Anemia, Confusion, blood Urea nitrogen, Respiratory rate).In the new ACUR rule, the present of anemia(+10) and confusion(+10), standardization blood urea nitrogen concentration >0.024(+10) and respiratory rate(/per minute), sum up the points assigned to the four characteristics, a total score will be calculated. And to improve the convenience, we also introduce the simplified ACR rule(Anemia, Confusion, Respiratory rate) which removes blood urea nitrogen and use the same calculate method.â…·) The classification of ACUR and ACRThere were no significant differences in inefficiency in each of the four risk classes among the six cohorts. In ACUR classification, inefficiency ranged from 42.1 to 50.0 percent for classâ… (No. of points: <20) patients, from 59.9 to 68.0 percent for classâ…¡(20-29), from 80.2 to 94.3 percent for classâ…¢(30-41), and 100.0 percent for classâ…£(>41). In ACR classification, inefficiency ranged from 36.7 to 54.5 percent for classâ… (No. of points: <20) patients, from 61.3 to 73.7 percent for classâ…¡(20-25), from 80.9 to 96.4 percent for classâ…¢(26-33), and from 97.9 to 100.0 percent for classâ…£(>33).â…¸) The evaluation of ACUR and ACRIn the receiver operating characteristic analysis, ACUR have a high quality to identify the inefficiency outcome: patients aged > or = 45 in FAG(Area Under the Curve [AUC], 0.745; 95% confidence interval [CI], 0.693-0.797), aged > or = 60 in FAG(AUC, 0.721; CI, 0.657-0.784), aged > or = 45 and < 60 in FAG(AUC, 0.809; CI, 0.726-0.893), aged > or = 45 in subgroup(AUC, 0.757; CI, 0.697-0.816), aged > or = 60 in subgroup(AUC, 0.746; CI, 0.676-0.816), aged > or = 45 and < 60 in subgroup(AUC, 0.806; CI, 0.701-0.911). ACR have a similar quality to identify the inefficiency outcome: patients aged > or = 45 in FAG(AUC, 0.747; CI, 0.695-0.800), aged > or = 60 in FAG(AUC, 0.726; CI, 0.661-0.790), aged > or = 45 and < 60 in FAG(AUC, 0.795; CI, 0.711-0.878), aged > or = 45 in subgroup(AUC, 0.769; CI, 0.708-0.829), aged > or = 60 in subgroup(AUC, 0.756; CI, 0.684-0.828), aged > or = 45 and < 60 in subgroup(AUC, 0.820; CI, 0.723-0.917). So ACUR and ACR have a high concordance (P=1.000) to predict the adverse outcome, but have significant differences betweenCURB65, CRB65, PSI, LRTI class (P=0.000).Limitations: In terms of variables, some of them can't reflect the true states of the patients. And of the rules, first, some patients have rare conditions which were not included but that clearly increase the likelihood of the poor outcome aren't suitable for the rules and can't be managed by it. Second, it can't identify patients with a low severity score who may require hospitalization for psychosocial and economic considerations. Third, the point of blood urea nitrogen will not be calculated easily. Conclusions: We propose the following recommendations cautiously after the analysis:â…°) There are 5 independent variables significantly associated with inefficacy outcome in elderly ( > or = 45 ) with CAP: increased respiratory rate and blood urea nitrogen, inadequate of hemoglobin, anemia, confusion.â…±) There are 4 independent variables significantly associated with inefficacy outcome in elderly ( > or = 60 ) with CAP: increased CRP and PSI, anemia, moist rales.â…²) There are 4 independent variables significantly associated with inefficacy outcome in patients aged > or = 45 and < 60 with CAP: increased neutrophil(%) and blood urea nitrogen, high school graduation level or less and severity.â…³) There are 7 independent variables significantly associated with inefficacy outcome in elderly ( > or = 45 ) with moderate and severe CAP: declined BMI, systolic pressure, hemoglobin, arterial partial pressure of oxygen, increased blood urea nitrogen, anemia, confusion.â…´) There are 7 independent variables significantly associated with inefficacy outcome in elderly ( > or = 60 ) with moderate and severe CAP: increased respiratory rate and PSI, declined systolic pressure and arterial partial pressure of oxygen, anemia, confusion and moist rales.â…µ) There are 1 independent variables significantly associated with inefficacy outcome in patients aged > or = 45 and < 60 with moderate and severe CAP: low albumin.â…¶) Although aging is not an independent variables significantly associated with inefficacy outcome, however, the clinical experience and the results of previous studies and the different confirmed factors in different age groups in the study give us an important indication that we can not easily rule out the correlation between aging and the poor outcome.â…·) ACUR, ACR rules are useful tools to predict inefficacy outcome in the different age groups and extent of an illness. ACR tool is easy to operate, but has less sensitive ability in discriminating the poor outcome in high-risk patients, which opposite to ACUR. |
PDF Full Text Request