Part of the midwives' role is to be vigilant towards detecting mental health problems that childbearing women may already have or develop. Of interest to this paper, is depression experienced across the childbearing spectrum. These are referred to as perinatal depression disorders (PDD) and include antenatal depression (AND) and postnatal depression (PND). PDD's can have significant consequences for women, partner, families and infants (Rahman et al, 2004), which includes reducing women's social participation, arousing sensitivity towards her newborn, causing infant malnutrition, physical illness and subsequent depressive episodes (Dix and Meuner, 2009; Josefsson and Sydsjö, 2007; Harvey et al, 2012). To date and in terms of research, AND has received far less research attention compared with PND, which is noteworthy because:
- Rates of AND and PND are similar with some evidence showing that the incidence of AND may be slightly higher than PND (Gavin et al, 2005)
- Evidence supports that AND may be a significant predictor for the woman developing PND (Beck, 2001)
- Research-based aetiology and interventions are limited for AND, compared with PND (Dennis and Dowswell, 2013; Chojenta et al, 2016; Morrell et al, 2016)
- Historical aetiological focus on PND has generally discounted the idea that PND may be an extension of pre-existing AND (Jomeen and Martin, 2008) yet in some cases this may be the case.
There is a confusing landscape of PDD's that can be diagnosed over the antenatal and postnatal period (Jomeen and Martin, 2008; Martin and Redshaw, 2018) which is surprising because depression is a commonly observed psychological phenomenon. However, recent studies illustrate how AND, akin to PND, is associated with adverse perinatal outcomes (Navaratne et al, 2016; Eastwood et al, 2017). In fact, studies have reported a higher prevalence for AND, from 6%–17% and PND, from 6%–12% (Leigh and Milgrom, 2008; Eastwood et al, 2011). A recent review by van Neil and Payne (2020) quotes that PDD affects 10%–20% of women during pregnancy and the postpartum period. Also, Sawyer et al (2010) identified rates of AND as 11.3% and PND to be 18.3%.
In comparison, the associated rates of antenatal and postnatal anxiety are 14.8% and 14% respectively, with anxiety traditionally thought to co-exist with depression (Zigmond and Snaith, 198; Christensen et al, 2020). Additional figures show that rates of severe depression are three times higher in the first postnatal year which is more than at any other time in a woman's life (World Health Organization, 2008). Consequently, it is important for midwives to screen for PDD's which across the continuum are labelled as AND and PND. Also worthy of consideration is whether or not PND is a specific type of depression that has its own distinctive aetiological pathway or instead is simply a continuation of pre-existing AND.
Aetiological models that underpin perinatal depression disorder
The first aspect of PDD that is important for midwives to acknowledge is that it can also profoundly affect husbands and/or partners (Cox, 2005; Psouni et al, 2017; Shaheen et al, 2019) with minimal attention paid to this position. Consequently, it is important for the midwife to have a toolkit of measures they can use to support suspicion that AND or PND may be present before referring the woman to a healthcare professional for formal assessment and diagnosis. Post-endorsement of a suspected diagnosis, the midwife can refer the woman or husband/partner for a formal assessment via the ‘mental health pathway’ outlined by their health board. Such action could forestall despair for a couple, with more studies needed to develop and measure effectiveness of different referral systems and treatment programmes.
The second aspect of PDD's that requires attention is the need to understand more about causal factors of AND and PND. For example, there is persuasive evidence to support that one cause of PND is routed in biology and hormones (Barry et al, 2015; Mah, 2016; Rogers et al, 2016) with a main contradiction to this argument being that males can also experience PND (Shaheen et al, 2019). The idea that husbands/partners can also experience PND is a concept that is antagonistic to biological aetiological models and use of pharmacological interventions to treat AND and PND (eg selective serotonin re-uptake inhibitors) (Milgrom et al, 2015; Ishikawa and Shiga, 2017; Molyneaux et al, 2018). Conflicting aetiological models about causes of PDD's have created an incomplete picture of cause, effect and appropriate treatments with more research required in this arena.
A third aspect of PDDs that requires attention is that currently the main screening focus of maternity care professionals is upon PND, with it recommended that equal attention be paid to diagnosis and treatment of AND. This recognition in disparity of attention is now just beginning to be acknowledged and yet has to fully infiltrate contemporary clinical guidelines (National Collaborating Centre for Mental Health, 2018). Although AND is as common as PND, mothers-to-be often miss out on proper treatment due to lack of midwife training. Midwives often do not receive sufficient formal training in AND and because some of the symptoms overlap with minor problems of pregnancy, such as tiredness and emotional instability, AND can be difficult to detect (Jomeen and Martin, 2008; Jomeen, 2009). Hence and of interest to the midwife, is what screening approach could be used to confirm need for referral of the woman for a formal diagnosis of AND or PND.
Screening for perinatal depression disorder
At present, there is confusion about what self-report measures a midwife could use to initially detect AND or PND with most measures validated for use post-childbirth. At present in the UK, the Edinburgh postnatal depression scale (EPDS) (Cox et al, 1987) is considered the ‘gold standard’ for initial detection of PND with it also validated for identifying AND (Murray and Cox, 1990). The midwife should note that cut-off-scores differ between antenatal and postnatal use of the EPDS (discussed later) (Murray and Cox, 1990).
It is important that the healthcare professional use the same valid and reliable psychometric self-report depression screening questionnaire with the same woman both antenatally and postnatally because:
This provides continuity of assessment, simply because the same questions are being asked at two separate observation points
- It allows a baseline to be recorded in the antenatal period against which a second observation point can be compared and improvement or deterioration of symptoms observed
- Consistency in scores across observation points (antenatal and postnatal), opens a debate about whether AND and PND are discrete conditions or a continuation of the same disorder.
These three points support the inherent fidelity of using the same valid and reliable psychometric self-report depression screening questionnaire before and after birth. Turning to the content of valid and reliable psychometric self-report depression screening questionnaires. There is an assumption that each scale is measuring the same thing which raises the concept of whether they are potentially interchangeable (Fried, 2017). It is important to note that switching scales could introduce sources of error which is another justification for why the same instrument should be used across longitudinal observation points with the same woman. Midwives can select from a variety of valid and reliable psychometric self-report depression screening questionnaires which can be used in combination to confirm or eliminate suspicion that a woman may have developed AND or PND.
1. ‘Whooley questions’
The two ‘Whooley questions’ (Whooley et al, 1997) are valid screening items asked by the midwife at the woman's first antenatal ‘booking visit’ in the UK (McGlone et al, 2016). The ‘Whooley questions’ have been shown to have acceptable utility in terms of sensitivity, specificity, and suitability as an initial screen for PDD (Arroll et al, 2005). The two ‘Whooley questions’ are:
- During the past month, have you often been bothered by feeling down, depressed or hopeless?
- During the last month, have you often been bothered by having little interest or pleasure in doing things?
The two ‘Whooley questions’ originate from the nine-item patient health questionnaire (PHQ-9) (Spitzer et al, 1994) and are answered by the woman with a straightforward ‘yes’ or ‘no’ response. When a positive screen is found, a third additional question is asked:
- Is this something with which you would like help?
The National Collaborating Centre for Mental Health (2018) guidelines recommend that the midwife conduct an initial ‘Whooley question’ screen and when depression is indicated, the woman is issued with a follow-up, valid and reliable psychometric self-report depression screening questionnaire. If total scores suggest that AND may be present, a post-interview assessment is undertaken by a mental health professional and an official diagnosis given. Each health board will have a guideline or protocol outlining their own mental health pathway which each midwife should be familiar with. Nonetheless, the official screening points may be limited, supporting the idea that midwives pay attention in-between when they are suspicious that a PDD may be present.
2. Edinburgh postnatal depression scale
The Edinburgh postnatal depression scale (EPDS) (Cox et al, 1987) is a 10-item easy-to-administer, valid and reliable psychometric self-report depression screening questionnaire. Scores above nine indicate ‘possible depression’ and beyond 12 ‘probable depression’. The EPDS has received many validations of effectiveness at screening for PDD (Hewitt et al, 2010) and has been translated and validated for use in at least 37 languages (Cox et al, 2014). It should be noted that most studies have focused upon use of the EPDS in the postnatal period with limited research conducted in the antenatal period.
One issue that the midwife needs to be aware of is that EPDS cut-off-scores for case classification differ between the antenatal and postnatal period. Murray and Cox (1990) recommend that a higher threshold for EPDS case classification is used in the antenatal period compared with the postnatal period (Gibson et al, 2009). For example, the cut-off-score of the original EPDS in the postnatal period is 12/13 (Cox et al, 1987). In comparison, in the antenatal period the cut-off score is 14/15 in the second and third trimesters of pregnancy (Murray and Cox, 1990; Adewuya et al, 2006).
Currently, there is no universal agreement about one single EPDS threshold cut-off-score for indicating diagnosis of AND or PND with this discrepancy causing differences in detection rates. Also, when considering the EPDS for use outside the UK, it is important to note that cut-off scores for countries and languages differ. To identify these language specific cut-off scores, the midwife can search the databases for country specific validation papers. It is important to acknowledge that substantially more is known about the screening utility of the EPDS in terms of its psychometric characteristics (ie validity, reliability and factor structure). Validity is largely based upon sensitivity and specificity analysis which has compared the EPDS against ‘gold standards’, such as clinical diagnosis according to the National Collaborating Centre for Mental Health (2018). Questions on the EPDS can be viewed at: https://www.fresno.ucsf.edu/pediatrics/downloads/edinburghscale.pdf
3. Nine-item patient health questionnaire
The nine-item patient health questionnaire (PHQ-9) (Spitzer et al, 1994) can be used to indicate AND or PND using a cut-off-score of 9/10. Please note that the PHQ-9 has different scoring approaches—it is important to be specific about which one should be used in any screening protocol or guideline. In relation to its screening ability, the PHQ-9 has similar sensitivity and specificity performance as the EPDS with this comparable performance suggesting that both instruments are suitable for screening for AND and/or PND (Zhong et al, 2014). Questions on the PHQ-9 can be viewed at: https://www.integration.samhsa.gov/images/res/PHQ%20-%20Questions.pdf
4. Beck depression inventory-11
Three versions of the Beck depression inventory (BDI) (Beck et al, 1961) have been developed with the BDI-II designed for use with individuals over the age of 13. The BDI-II is a revision of the original BDI (Beck, 1996a) which was adapted in response to the DSM-4 changing its diagnostic criteria for major depressive disorder. The BDI-II consists of 21-items that the woman scores from 0–3 with higher total scores indicating more severe depressive symptoms. Cut-off-scores of the BDI-II rate 0–13 minimal depression, 14–19 mild depression, 20–28 moderate depression, and 29–63 severe depression. Questions on the BDI-11 can be viewed at: https://psychologicalprofessional.com/wp-content/uploads/2017/07/Becks-Depression-Inventory-BDI-II.pdf
5. Hospital anxiety and depression scale
The hospital anxiety and depression scale (HADS) determines levels of both anxiety and depression with anxiety considered to co-exist with depression (Zigmond and Snaith, 1983). The HADS consists of 14-items (seven measure anxiety; seven measure depression). Items are rated on a four-point scale from 0–3, creating a maximum achievable score of 21 for each sub-scale (42 whole scale total). Individual sub-scores of 11+ indicate significant psychiatric comorbidity, scores of 8–10 signify presence of condition, and seven or less indicate normalcy (Hermann, 1997).
There have been many explorations into the HADS measurement characteristics which have shown the scale to be a reliable and valid screening instrument (Martin and Thompson, 2000; Norton et al, 2013; Christensen et al, 2020) which is suitable for use across the childbearing spectrum (Karimova and Martin, 2003; Jomeen and Martin, 2004; Waqas et al, 2019). Consistent with the EPDS, the HADS has no body related (somatic) items that affect mood which are caused by physiological changes that occur during the antenatal and postnatal period (eg morning sickness, backache, urinary frequency, anaemia, weight gain, varicosities etc). Questions on the HADS can be viewed at: https://www.svri.org/sites/default/files/attachments/2016-01-13/HADS.pdf
6. Hamilton depression rating scale
The Hamilton depression rating scale (HDRS) is a multiple-item questionnaire used to indicate depression (Hamilton, 1967). The original 1960 version contains Hamilton depression rating scale 17-items (HDRS-17) but since then four additional questions have been added to provide extra clinical information (Hamilton, 1980). Each item is scored by the respondent on a 3–5 point Likert scale with items asking about mood, feelings of guilt, suicide ideation, insomnia, agitation, anxiety, weight loss and somatic symptoms. Questions on the HADS can be viewed at: https://dcf.psychiatry.ufl.edu/files/2011/05/HAMILTON-DEPRESSION.pdf
Limitations
Having described six options that can be used to indicate that a woman has AND or PND, it is important for the midwife to understand limitations of using such valid and reliable psychometric, self-report depression screening questionnaires. First, responses can be easily exaggerated or minimised by the person completing items on the scale. Second, the way the midwife administers the scale can impact upon the final total score. Third, when the woman is asked to complete the questionnaire with the midwife present, social expectations may prompt responses that differ to competing the scale in private. Fourth, valid and reliable psychometric, self-report depression screening questionnaires are usually copyrighted, hence only links to the scales have been provided in text for the reader to view question content. Some scales cost money to use and others are free. Check with management whether they already have an annual or ongoing license for use of a chosen scale. Fifth, any valid and reliable psychometric, self-report depression screening questionnaire is designed to be a checking device as opposed to a diagnostic tool. Consequently, any indicative diagnosis of PDD made by the midwife must be followed up with an interview and diagnosis made by a trained mental health professional.
There is also the issue of exchangeability of scales which involves researchers checking for instrument usefulness by comparing how closely one scale agrees with another. For example, the BDI-II has been positively correlated with the HDRS (Pearson r=0.71) which is a good agreement (Beck et al, 1996a). In terms of reliability, the BDI-II has also been shown to have a high one-week test-retest reliability (Pearson r=0.93) which suggests that the scale is not overly sensitive to daily variations in mood (Beck et al, 1996b). These sorts of inventories have been conducted for all the aforementioned scales with references available for retrieval in the databases.
Discussion
Screening with valid and reliable psychometric, self-report depression screening questionnaires is fundamentally a cost-effective approach to initially indicate whether a woman has AND or PND. The large amount of choice over which self-report measures a midwife should use is a conundrum with it noted that apart from the EPDS, the other aforementioned scales were developed to detect general depression as opposed to screening specifically for AND or PND. Clearly, more studies are required to evaluate effectiveness and accuracy of these scales in an antenatal and/or postnatal context. It is also important to identify optimum characteristics of how best to use a single scale or combination of valid and reliable psychometric, self-report depression screening questionnaires in a toolkit. Also of importance is to consistently issue the same scale at set observation points across an individual woman's childbearing spectrum.
In relation to validation of effectiveness of valid and reliable psychometric, self-report depression screening questionnaires, many studies conducted have not been cross-sectional in terms of giving the same scale across set observational time points. One study that supports cross-sectional use of the EPDS was conducted by Martin and Redshaw (2018). Although scores changed across two observational time points (three and six months), the underlying structure of the EPDS was found to be consistent in terms of its psychometric properties (Martin and Redshaw, 2018).
Selecting an appropriate toolkit
Conciseness is key when considering which valid and reliable psychometric, self-report depression screening questionnaires to include in a toolkit for the purpose of initially indicating whether or not a woman may have AND or PND. During process of selection for inclusion into clinical protocols and guidelines, it is important to state the cut-off-thresholds for AND or PND. The National Collaborating Centre for Mental Health (2018) states that the main criteria when selecting a valid and reliable psychometric, self-report depression screening questionnaire is that the scale should consist of 12-items or less.
This restriction is due to time costs and sustaining interest of the woman in terms of full completion. This 12-item restriction promotes that the only suitable scales for use after asking the two ‘Whooley questions’ is the 10-item EPDS or the nine-item PHQ-9. This recommended restriction in question numbers is unfortunate, precisely because there are several valid and reliable psychometric, self-report depression screening questionnaires that have dependable utility and which consist of more than the allocated quota of questions (ie the 21-item BDI-11; 14-item HADS; 17-item HDRS). Note that the 14-item HADS consists of two seven-item scales (seven-items indicate anxiety and seven-items indicate depression), with the seven-items that check for depression possibly being used independently.
In practical terms, these prohibited scales often take similar amount of time for the woman to complete. For instance, the 14-item HADS has had many explorations into its measurement characteristics (Martin and Thompson, 2000; Norton et al, 2013; Christensen et al, 2020) which includes the context of pregnancy and the postnatal period (Karimova and Martin, 2003; Jomeen and Martin, 2004; Waqas et al, 2019). When considering potential utility of the 14-item HADS in a perinatal context, it is helpful to know that it consists of two discrete sub-scales of anxiety (seven-items) and depression (seven-items). The HADS has four more questions than the 10-item EPDS (14 versus 10 respectively). Yet, the advantage the HADS has over the EPDS is that it screens for both depression and anxiety.
This added screening for anxiety is useful specifically because anxiety is reported to co-exist with depression (National Collaborating Centre for Mental Health, 2018). In addition to detecting depression, the further benefit of detecting anxiety using the HADS with an additional four-item effort is its cost-effectiveness in terms of time and resources. Overall and in general terms, the HADS is by far the most frequently used screening measure for identifying both depression and anxiety in a variety of contexts (Christensen et al, 2020). Also, the HADS has a substantial evidence-base to verify its use within a broad range of clinical groups (Martin and Thompson, 2000; Karimova and Martin, 2003; Jomeen and Martin, 2004; Norton et al, 2013; Waqas et al, 2019Christensen et al, 2020).
Broader concepts of validity, such as assessing factor structure, are generally not considered important when combining self-report measures into toolkits, guidelines and protocols. This oversight brings with it a few challenges worthy of consideration. First, the ‘Whooley questions’ are valid to use for a quick initial mental health screen. Post assimilating a confirmatory response to the ‘Whooley questions’, clinical guidelines usually advise follow-ups using either the EPDS or PHQ-9. However, the choice of whether to use either of these self-report measures is arbitrary and rests upon clinician preference.
This quandary about what self-report measures to use to detect AND and PND is based upon belief that both the EPDS or PHQ-9 are measuring the same which conceptually makes them essentially interchangeable scales. In relation to this concept of exchangeability, some literature reports that there is little overlap in the content of frequently used valid and reliable psychometric, self-report depression screening questionnaires while, in contrast, there are reports that exchanging measures produces errors. What this means to the midwife and researchers is that when comparing across longitudinal observation points over the antenatal and postnatal spectrum, the same valid and reliable psychometric self-report depression screening questionnaire should be used. It is also important that further research in this area of measurement error and exchangeability of valid and reliable psychometric, self-report depression screening questionnaires is both required and encouraged.
Conclusion
Valid and reliable psychometric self-report depression screening questionnaires can be very useful for indicating women who are struggling with depression during the antenatal or postnatal period. Such scales can help midwives screen and refer women for appropriate treatment for their depression from mental health professionals. It is a given that any screening questionnaire that is embedded into a clinical guideline or protocol should be statistically valid and reliable with this belief, based upon a very narrow definition of sensitivity and specificity.
This paper has raised important issues for midwives that encompass identification of depression in childbearing women across the perinatal spectrum. What has been identified is that midwives screening for AND is as important as screening for PND, simply because rates of both are similar. Prior research emphasis has traditionally been focused upon detecting PND. What is important is for midwives to develop a toolkit of valid and reliable psychometric, self-report depression screening questionnaires that they can use in conjunction with guidelines and protocols. It is also important to make these toolkits available to women free of charge in pamphlets, online and embedded into mobile applications so they can self-screen and optionally refer themselves.
Key points
- Midwives screening for antenatal depression (AND) is as important as screening for postnatal depression (PND)
- AND rates are comparable to PND rates yet emphasis of perinatal screening is conventionally upon PND
- AND can be successfully and cost-effectively identified by the midwife, through using valid and reliable psychometric self-report depression screening questionnaires eg the ‘Whooley Questions’, Edinburgh postnatal depression scale, nine-item patient health questionnaire, Beck depression inventory, hospital anxiety and depression scale, Hamilton depression rating scale
- It is important for midwives to have a toolkit that can be given to women at clinics, in pamphlets, online or embedded into mobile apps for self-detection of perinatal depression disorders
- Further research is needed to examine the degree of overlap of questionnaire items between measures in order to determine whether they have acceptable commonality in assessing a consistent conceptual domain of depression
CPD reflective questions
- What aspects of clinical midwifery practice could be developed to improve detection, referral and treatment of antenatal depression (AND) and postnatal depression (PND)?
- Consider which scale you would add to your ‘toolkit’ for detecting AND or PND.
- What education should be provided to midwives to enhance their ability to provide care for women experiencing AND or PND?
- What research is needed to develop standards of maternity care provided to women, partners and families who are experiencing perinatal mental health complications?