his/md/门诊就诊记录SQL优化建议.md

门诊就诊记录SQL查询优化建议

当前查询分析

主要查询表

SELECT
    enc.id as encounterId,
    pt.name,
    pt.id_card,
    pt.bus_no as patientBusNo,
    enc.bus_no as encounterBusNo,
    pt.gender_enum,
    pt.phone,
    enc.create_time as encounterTime,
    enc.status_enum as subjectStatusEnum,
    org.name as organizationName,
    prac.name as doctorName
FROM adm_encounter AS enc
LEFT JOIN adm_organization AS org ON enc.organization_id = org.ID AND org.delete_flag = '0'
LEFT JOIN adm_encounter_participant AS ep
    ON enc.ID = ep.encounter_id AND ep.type_code = #{participantType} AND ep.delete_flag = '0'
LEFT JOIN adm_practitioner AS prac ON ep.practitioner_id = prac.ID AND prac.delete_flag = '0'
LEFT JOIN adm_patient AS pt ON enc.patient_id = pt.ID AND pt.delete_flag = '0'

常见查询条件

1. enc.delete_flag = '0'
2. enc.tenant_id = ?
3. pt.name LIKE ?
4. pt.id_card LIKE ?
5. pt.bus_no LIKE ?
6. enc.bus_no LIKE ?
7. pt.gender_enum = ?
8. enc.status_enum = ?
9. prac.name LIKE ?
10. pt.phone LIKE ?
11. enc.create_time BETWEEN ? AND ?

索引优化建议

1. adm_encounter 表索引

-- 复合索引：提高查询性能
CREATE INDEX idx_encounter_tenant_delete_status ON adm_encounter(tenant_id, delete_flag, status_enum);

-- 时间范围查询索引
CREATE INDEX idx_encounter_create_time ON adm_encounter(create_time);

-- 业务编号查询索引
CREATE INDEX idx_encounter_bus_no ON adm_encounter(bus_no);

-- 患者ID关联索引
CREATE INDEX idx_encounter_patient_id ON adm_encounter(patient_id);

2. adm_patient 表索引

-- 姓名模糊查询索引
CREATE INDEX idx_patient_name ON adm_patient(name);

-- 身份证号查询索引
CREATE INDEX idx_patient_id_card ON adm_patient(id_card);

-- 业务编号查询索引
CREATE INDEX idx_patient_bus_no ON adm_patient(bus_no);

-- 电话查询索引
CREATE INDEX idx_patient_phone ON adm_patient(phone);

-- 复合索引：常用查询条件
CREATE INDEX idx_patient_delete_gender ON adm_patient(delete_flag, gender_enum);

3. adm_encounter_participant 表索引

-- 复合索引：提高连接性能
CREATE INDEX idx_ep_encounter_type ON adm_encounter_participant(encounter_id, type_code, delete_flag);

-- 参与者ID索引
CREATE INDEX idx_ep_practitioner ON adm_encounter_participant(practitioner_id);

4. adm_practitioner 表索引

-- 姓名查询索引
CREATE INDEX idx_practitioner_name ON adm_practitioner(name);

-- 复合索引：常用查询条件
CREATE INDEX idx_practitioner_delete_tenant ON adm_practitioner(delete_flag, tenant_id);

5. adm_organization 表索引

-- 主键关联索引
CREATE INDEX idx_organization_id_delete ON adm_organization(id, delete_flag);

查询优化建议

1. 添加查询统计信息收集

-- 定期分析表统计信息
ANALYZE TABLE adm_encounter;
ANALYZE TABLE adm_patient;
ANALYZE TABLE adm_encounter_participant;
ANALYZE TABLE adm_practitioner;
ANALYZE TABLE adm_organization;

2. 考虑分区表（针对大数据量）

如果 adm_encounter 表数据量超过100万条，考虑按时间分区：

-- 按月分区
PARTITION BY RANGE (YEAR(create_time) * 100 + MONTH(create_time))
(
    PARTITION p202501 VALUES LESS THAN (202501),
    PARTITION p202502 VALUES LESS THAN (202502),
    -- ... 更多分区
);

3. 添加覆盖索引（Covering Index）

对于常用查询字段，创建覆盖索引避免回表：

CREATE INDEX idx_encounter_cover ON adm_encounter(
    tenant_id, delete_flag, create_time, 
    status_enum, bus_no, patient_id
) INCLUDE (organization_id);

执行计划检查

建议定期检查查询执行计划：

EXPLAIN ANALYZE
SELECT -- 完整查询语句
FROM adm_encounter AS enc
-- ... 连接条件
WHERE enc.delete_flag = '0'
  AND enc.tenant_id = 1
  -- ... 其他条件
ORDER BY enc.create_time DESC;

监控建议

1. 慢查询监控：监控执行时间超过1秒的查询
2. 索引使用监控：定期检查未使用的索引
3. 表空间监控：监控表增长和碎片情况
4. 连接性能监控：监控JOIN操作的性能

实施步骤

1. 在测试环境创建建议的索引
2. 执行查询性能测试
3. 分析执行计划，确认索引有效性
4. 在生产环境非高峰期创建索引
5. 监控生产环境性能变化
6. 定期维护和优化索引