Commentary on Hung C-H (2006) Revalidation of the postpartum stress scale. Journal of Clinical Nursing 15, 718–725


Andrew Symon, School of Nursing & Midwifery, University of Dundee, Dundee, UK. Telephone: 01382 496671,

This article by Hung (2006) reports the re-testing of a postpartum stress scale which was first devised in 1993. Hung notes that changes in Taiwan since the scale was originally developed indicated a need to verify its content. Despite what seemed to be thorough testing of the original tool, one double- and one triple-barrelled question still managed to make it through to the final version. Although this is not mentioned in this paper, a further stage in the testing of this scale (involving an 85-item version) appears to have been carried out with 861 postnatal women from several hospitals and clinics in Taiwan (Hung 2005).

The postnatal period is a time of great adaptation and change, but is also said to be a period that receives comparatively little attention in terms of practice, teaching and research (Albers 2000). Studies in the West have often focused around issues of physical morbidity, with a growing recognition that this can be more complex than was once thought (Glazener 1997). There has also been recent interest in looking beyond physical and emotional morbidity and examining postnatal quality of life (Symon et al. 2003). Such approaches acknowledge that the mother is an individual with specific concerns, and will be affected by her particular circumstances – what Hopkins (1992, p. 1) calls the ‘social reality of the situation’. Hung notes that, in Taiwan, there are certain specific cultural issues which make the first month following childbirth a potentially stressful time for mothers. Amongst these are cited the traditional ritual of seclusion and restriction of physical activity, the naming of the baby and the baby's facial shape, which is thought to be crucial to his/her fortune.

In Hung's study, which took place in the southern Taiwanese city of Kaohsiung, 505 and 518 postnatal women were surveyed at, respectively, one and five weeks following childbirth. They were asked to rate responses to 66 items on a Likert scale, following which their responses were subjected to factor analysis. Of the original 66 items, 54 were found to have salient loading at one week and 38 were found to have salient loading at five weeks. It was not apparent from the paper whether this meant that two versions of the stress scale were therefore appropriate at the two specified time points.

Hung reports that this process supported the identification of three principal dimensions that relate to postnatal stress: ‘maternal role attainment’, ‘lack of social support’ and ‘negative body changes’. These three dimensions were found to have high Cronbach alpha scores (0·76–0·90) when the first and fifth week scores were compared.

Given that almost half of Hung's sample (49·6%) were primiparous, I was not sure how their responses to the item ‘lack of time to care for my other children’ would be interpreted. The allocation of certain items to the three dimensions was, at times, puzzling. At week 1 three references to sexual intercourse (‘feeling uncomfortable during sexual intercourse’, ‘my sexual life because of stretching of the vagina’, and ‘decreased frequency of sexual intercourse’) were all listed under ‘lack of social support’. Hung notes that one of the traditional Chinese medicine rituals for women is to avoid sexual intercourse in the first postnatal month, so questions about sex appear anomalous, particularly at one week (Alder 1989). In addition, the first two items were relocated to the dimension ‘negative body changes’ at the five-week stage. The third example (‘decreased frequency of sexual intercourse’) was not one of the 38 items that achieved salient loading at five weeks. Similarly, ‘neglecting my husband's care’ and ‘increased family expenses’, which had been listed under ‘lack of social support’ at week 1, were also listed under ‘negative body changes’ at week 5. Another item (‘poor contractions of the uterus’) was classified under ‘maternal role attainment’ (alongside other items concerning feeding, dressing and raising the baby), which was not an immediately obvious choice. The rationale for how these items were originally classified or for relocating some of them was not clear.

Hung notes that this study involved a ‘representative sample of Taiwanese women’. No doubt because of understandable ethical considerations, the testing of this assessment tool was restricted to certain groups of mothers: those who had had a single, healthy, full-term baby, without complications; and those who had had no major postnatal complications or underlying medical problems. While understandable, this would appear to limit the confidence with which the tool can be said to be generally applicable. Indeed, it would be particularly interesting to identify the stress levels of those who did not satisfy the inclusion criteria, as they may be the very women who require the most assistance and support.

In terms of the representativeness of the sample, it was interesting to see that 45% of the women had had a Caesarean section. At face value this is very high. In fact in Hung's other paper (Hung 2005) concerning testing of this scale, it is noted that 44% of the 861 women had had a vaginal delivery, indicating a Caesarean section rate of 56%. There is concern within the UK and the USA about the Caesarean rate, where levels appear to fluctuate between 20% and 25% in many units. Taiwan, like the UK and the USA, has seen a rapid increase in the incidence of Caesareans over the last four decades, although the most recent report I could identify notes that the Taiwanese rate is around 34% (Lo 2003), which would bring into question this sample's representativeness.

There are interesting issues to do with cultural sensitivity in a study such as this. Certainly, it is important to use assessment tools that relate to particular circumstances, whether these are national or individual. In Taiwan the cultural norm is for postnatal women to be confined to the home for a month, with much of the daily household activities carried out by her mother-in-law, or now, increasingly, her own mother. Hung notes that this change has arisen because of ‘the demise of the extended kinship family’. However, the claim that being ‘bothered by feeling that my life is confined’ is a ‘Taiwanese-specific sociocultural’ issue was not convincing. While there certainly appear to be particular issues to do with Taiwanese postnatal customs, perceived isolation has been described elsewhere in immigrants (Barclay & Kent 1998), high-risk adolescent mothers (Williams & Vines 1999) and women who are at risk of postnatal depression (Forman et al. 2000, Russell 2005).

There are other assessment tools that measure adjustment in the postnatal period (e.g. Kumar et al. 1984). Hung notes that cultural norms vary between countries and rightly states that assessment tools must take account of these. The very specific references to Taiwanese cultural practices mean that while this stress scale may be applicable to Taiwanese mothers (despite the caveats I have noted above) its applicability outside Taiwan is questionable. However, it is helpful to be informed of the different ways in which the postnatal period is viewed across the world. That said, the crux for any assessment tool is that it leads to targeted treatment or improved case management, and in this respect the efficacy of this stress scale has yet to be reported.