Question

以下查询几乎适用于所有方面，但它会返回遇到的总数（pe.enc_id）。我试图在2016年找到总访问量。

select distinct p.person_nbr, p.last_name, p.first_name, p.race, count(pe.enc_id) as count_of_encounters 
from patient_encounter pe
    inner join person p on pe.person_id = p.person_id
    inner join patient_diagnosis pd on pd.enc_id = pe.enc_id
WHERE (p.date_of_birth between '19420101' and '19981231' and pe.enc_timestamp >= '20160101' AND pe.enc_timestamp  < '20170101')
    and pd.person_id in (select distinct p.person_id from person p  inner join patient_diagnosis pd on p.person_id = pd.person_id
where (pd.diagnosis_code_id  like '250%' or pd.diagnosis_code_id like '648.0%' 
    or pd.diagnosis_code_id  like '357.2' or pd.diagnosis_code_id like '362.0[0-7]' or pd.diagnosis_code_id  like '366.41' or pd.diagnosis_code_id like 'E1[0-3].%' 
    or pd.diagnosis_code_id  like 'O24.[0-4]0%' or pd.diagnosis_code_id like 'O24.[4-9][2-9]') )
group by p.person_nbr, p.last_name, p.first_name, p.race
having count(pe.enc_id) >= 2

Answer 1

您的查询确实看起来是在日期过滤。我有一些建议：

您无需select distinct
您的子查询过于复杂。
事实上，我认为没有必要。

您似乎正在寻找具有特定诊断的患者，然后计算诊断数量。您可以通过扩展having子句在外部聚合中完成所有操作：

所以：

select p.person_nbr, p.last_name, p.first_name, p.race,
       count(pe.enc_id) as count_of_encounters 
from patient_encounter pe inner join
     person p
     on pe.person_id = p.person_id inner join
     patient_diagnosis pd
     on pd.enc_id = pe.enc_id
where p.date_of_birth between '19420101' and '19981231' and
      pe.enc_timestamp >= '20160101' AND pe.enc_timestamp  < '20170101') 
group by p.person_nbr, p.last_name, p.first_name, p.race
having count(pe.enc_id) >= 2 and
       sum(case when pd.diagnosis_code_id  like '250%' or pd.diagnosis_code_id like '648.0%' 
    or pd.diagnosis_code_id  like '357.2' or pd.diagnosis_code_id like '362.0[0-7]' or pd.diagnosis_code_id  like '366.41' or pd.diagnosis_code_id like 'E1[0-3].%' 
    or pd.diagnosis_code_id  like 'O24.[0-4]0%' or pd.diagnosis_code_id like 'O24.[4-9][2-9]' then 1 else 0 end) > 0;

用于查找日期范围内的访问次数的SQL查询

1 个答案: