Why I got incorrect number of All obs after PROC PSMATCH procedure?

[vivianwang]

Hello SAS Community,

I'm doing 1:1 propensity score matching procedure

The data set contains 47366 samples (treatment=30694/ control=16672)

After PROC PSMATCH procedure, I got a table as below

The number of All Obs (Treatment) and All Obs (Control) is different from the original data set, 70 of the total sample were lost. 

And I also got a warning: Some treated units do not have matched controls because there are not enough available controls for these treated units.

Is this problem related to the warning? How do I solve it?

proc psmatch data=B.HLP_ID_COMED  region=treated; 
class SEX statin c_htn c_dm c_os c_mg c_go c_as c_sch c_emd c_nd c_dp c_thy c_gl m_antihi m_bb m_diur m_ssri m_anticho  m_ppi m_glau ;
Psmodel statin (Treated='1')= AGE SEX c_htn c_dm c_os c_mg c_go c_as c_sch c_emd c_nd c_dp c_thy c_gl m_antihi m_bb m_diur m_ssri m_anticho  m_ppi m_glau; 
match method=greedy( k=1) exact=(SEX)  caliper=0.2 ; 
assess lps var=(AGE SEX c_htn c_dm c_os c_mg c_go c_as c_sch c_emd c_nd c_dp c_thy c_gl m_antihi m_bb m_diur m_ssri m_anticho  m_ppi m_glau)/plots=ALL  weight=none; 
output out(obs=match)=OUT_PSM matchid=_MatchID; run;

[Oligolas]

Hi,

did you check if you have missings in your treatment variable?

[vivianwang]

Thank you for your reply 

I’ve checked the sample and  there is no missing data.

Is there other explanations for the 70 lost from the sample?

Thank you!