We read with interest the recent article by Venkatesh et al regarding the implementation of routine antenatal and postnatal screening for depression for women in the perinatal period in Massachusetts, using the Edinburgh Postnatal Depression Scale (EPDS) (Cox et al ). We would like to make some observations about this study and its wider implications for other services considering the use of this self-report scale.

• 1.
  

  As the authors correctly state, many studies use a cut-off score of ≥12 on the EPDS. This, however, is often due to an error in these studies reporting the validation study by Cox et al, in which ≥13 was in fact the validated cut-off score for English-speaking women in the postpartum period. This error has previously been shown to be frequent, and that a difference of just 1 point in the cut-off score does indeed have a significant impact on findings and is not just trivial. Unfortunately this mistake is often still being made, and thus gives the impression that ≥12 is the validated cut-off score for English-speaking postpartum women, when it is not.

Indeed, it is somewhat concerning that the authors themselves have made a similar type of error. They cite Murray and Carothers as evidence for studies using ≥12 as the cut-off score, when that article does not in fact make any recommendation for this score, but highlights (without making any specific score recommendations) the scores of ≥11 and ≥13, and discusses that Cox et al found ≥13 to be the optimal cut-off score.

• 2.
  

  Venkatesh et al used the same EPDS cut-off score for the antenatal and postnatal periods, without commenting on the research showing that different cut-off scores are required for the 2 time periods. In addition it is important that services know that there is considerable evidence that different cut-off scores on the EPDS are required for women (and men) from different cultures.

  
  
• 3.
  

  While Venkatesh et al report that the EPDS data were successfully entered into the electronic medical record, this does not necessarily mean that the score that was entered was in fact correct. Services need to be aware that the accuracy of clinicians scoring the EPDS has been shown to be very poor, even among those who provide training in the use of this scale.