Original article
Psychometric evaluation of health-related quality of life measures in women after different types of delivery

https://doi.org/10.1016/j.jpsychores.2007.06.003Get rights and content

Abstract

Objective

We examined the psychometric properties of three internationally established measures for health-related quality of life (HRQoL) in women after vaginal delivery (VD), elective cesarean section (CS), and emergency CS and the relationship of HRQoL scores with blood loss after delivery.

Methods

This is a prospective longitudinal study. One hundred forty-one consecutive patients (71 after VD, 36 after elective CS, and 34 after emergency CS) were enrolled in two university hospitals and one general hospital from June 2003 to March 2004. Women completed the Multidimensional Fatigue Inventory (MFI) and the EQ-5D classification of own health between 12 and 24 h after VD or between 24 and 48 h after CS. Subsequent assessments, additionally including the Short Form 36 (SF-36), were made 1, 3, and 6 weeks after delivery. We analyzed feasibility (response, completion time, reported difficulties, item nonresponse), reliability (Cronbach's α), discriminative validity between groups by type of delivery, and responsiveness over time (Wilcoxon's signed rank tests and effect sizes).

Results

The MFI, SF-36, and EQ-5D proved to be highly feasible and reliable (α>.7 for all scales of MFI and SF-36). The measures were able to discriminate between groups by mode of delivery and to detect moderate recovery in physical and small recovery in mental status over time in the first 6 weeks after delivery. The suboptimal total questionnaire response of 60% after 6 weeks was attributable to low response among women of non-Dutch ethnic origin. The significant correlation between Hb level and mean physical HRQoL scores found at T=0 disappeared 1 week postpartum.

Conclusion

The combination of MFI, SF-36, and EQ-5D showed good psychometric performance and is a good choice to measure HRQoL after delivery in scientific studies. Development of a shorter set is needed for use in routine clinical practice.

Introduction

Patient-reported outcomes, including health-related quality of life (HRQoL), have increasingly been incorporated in the evaluation of medical technology assessments. HRQoL is often defined as patient-reported functioning and well-being in the physical, psychological, and social domains. Generally, there are three types of HRQoL measures: (a) generic measures, intended for use both in general population surveys and in studies of patients with diverse health conditions, allowing for comparison of HRQoL scores across disease stages and diagnostic groups; (b) condition-specific measures, developed for use among specific patient population (e.g., cancer, diabetes); and (c) domain-specific measures, for measurement of specific symptoms (e.g., fatigue, pain). It is common practice to combine condition-specific and/or domain-specific measures with generic measures. The feasibility and other psychometric properties of HRQoL measures, however, may differ between populations.

The birth of a child has a major impact on the new mother. Her HRQoL is influenced by a variety of medical, psychological, social, and obstetric factors (e.g., her age, general physical health during pregnancy, length of labor, maternal expectations, beliefs, attachment, and mood). Traditionally, the postpartum period is considered to be 6 weeks [1]. Longitudinal studies measuring HRQoL postpartum show serious physical and emotional problems in more than 50% 1 year postpartum [2], [3]. Some of these symptoms are still present more than 12 months postpartum [4]. HRQoL in patients after vaginal delivery (VD) was found to be significantly different from HRQoL in women after cesarean section (CS) [3]. With a prevalence of 60–70%, fatigue is the most frequent symptom after 1 year [2], [4]. To date, studies about fatigue and HRQoL in the immediate postpartum period are lacking.

Blood loss during delivery is one of the most frequently occurring delivery-related complications that may have consequences for maternal HRQoL in the immediate postpartum period and thereafter, for example, because of excessive fatigue [5]. Red blood cell (RBC) transfusion is one of the primary treatments for postpartum anemia. In general, the medical ground for an RBC transfusion is based on Hb triggers [6]. The aim of RBC transfusions postpartum is, however, not to increase Hb values but to improve HRQoL. With the exception of extremely low Hb levels, there is no medical need to restore a low Hb level with RBC transfusions directly postpartum. Moreover, a blood transfusion is a generally safe procedure but it is not completely without risks [7], [8].

However, HRQoL data are currently not applied in clinical decision-making process regarding blood transfusions in the postpartum period. Therefore, we started a prospective randomized study comparing a novel RBC transfusion policy after delivery and usual care—the WOMB study (Well-being of Obstetric Patients on Minimal Blood Transfusions; ClinicalTrials.gov identifier: NCT00335023). HRQoL is the primary outcome measure in the WOMB study. Preceding the start of the WOMB study, we tested the appropriateness of various HRQoL measures, including the domain-specific Multidimensional Fatigue Inventory (MFI) for fatigue and two generic measures [Short Form 36 (SF-36) and the EQ-5D classification of own health], among women after VD, elective CS, and emergency CS. This study aimed to assess, in a clinical obstetrical setting, the following:

  • 1.

    the feasibility of the MFI, SF-36, and EQ-5D (indicators: response rate, missing/nonunique answers, reported difficulties, and completion time);

  • 2.

    the score distribution of the MFI, SF-36, and EQ-5D (mean scores, standard deviations, presence of floor and ceiling effects);

  • 3.

    the reliability of the scales of the MFI and SF-36 (internal consistency);

  • 4.

    the discriminative ability of the MFI, SF-36, and EQ-5D rating of own health between groups of women by type of delivery 1 week after delivery;

  • 5.

    the responsiveness over time of the MFI, SF-36, and EQ-5D rating of own health (differences between mean scores at T=0 and T=6 weeks and effect sizes); and

  • 6.

    the relationship between blood loss and MFI fatigue scores.

Section snippets

Patients

In the period June 2003 to March 2004, 71 consecutive patients after VD and 70 patients after CS were included in two university hospitals and in one general hospital in Rotterdam and Leiden (the Netherlands). Patients undergoing CS were post hoc divided into those undergoing an elective CS (n=36) and those undergoing an emergency CS (n=34). Indications for elective CS were breech and transverse presentation, two or more previous CS, previous anorectal surgery, maternal condition, and a

Feasibility

A total of 55% of the patients returned all HRQoL measures, 16% returned three questionnaires, 9% returned two questionnaires, and 20% returned only the first questionnaire. Of the Dutch ethnic patients, 70% returned all questionnaires, whereas only 44% of the non-Dutch ethnic patients returned all questionnaires. For all different time points combined, the mean amount of missing values (range) of the returned HRQoL measures was 1.8 (0–4) for the MFI, 1.1 (0–3) for EQ-5D classification,

Discussion

This evaluation of the international standard HRQoL measures MFI, EQ-5D, and SF-36 established the feasibility, reliability, and validity of these measures in a clinical obstetric setting. This study supports the discriminative ability of these measures by type of delivery and provides reference (norm) scores for patients after delivery.

Different HRQoL measures are used in studies with patients after delivery [25], [26], [27], [28]. However, no psychometric evaluation of HRQoL measures has been

References (34)

  • MJ Saurel-Cubizolles et al.

    Women's health after childbirth: a longitudinal study in France and Italy

    BJOG

    (2000)
  • TL Simon et al.

    Practice parameter for the use of red blood cell transfusions

    Arch Pathol Lab Med

    (1998)
  • EAM Beckers et al.

    Reports of transfusion incidents: experiences from the first year of hemovigilance in the region of the former ZWN (South West Netherlands) blood bank in Rotterdam (article in Dutch)

    Ned Tijdschr Geneeskd

    (2003)
  • L Williamson et al.

    The Serious Hazards of Transfusion (SHOT) Initiative: the UK approach to haemovigilance

    Vox Sang

    (2000)
  • AJG Jansen et al.

    New insights into fatigue and health-related quality of life after delivery

    Acta Obstet Gynecol Scand

    (2007)
  • CBS Index...
  • EMA Smets et al.

    Application of the Multidimensional Fatigue Inventory (MFI-20) in cancer patients receiving radiotherapy

    Br J Cancer

    (1996)
  • Cited by (64)

    • Patient-reported outcome measures evaluating postpartum maternal health and well-being: a systematic review and evaluation of measurement properties

      2022, American Journal of Obstetrics and Gynecology MFM
      Citation Excerpt :

      The domain of role transition was recommended by 3 “A”-recommended tools: PQOL, BIMF, and PPSS-S. Incontinence and pain with intercourse were not covered by any “A”-rated tool. There were 3 tools in our review (covering fatigue and social support in isolation) that did not have sufficient content validity and were awarded a “C” recommendation; these were the fatigue assessment scale (FAS),28 MFI,42 and MOS-SSS.21 The tools recommended as “B” require further evaluation of their quality before recommendation for use.

    View all citing articles on Scopus
    View full text