SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

Objective  To determine the impact of a national strategy to promote implementation of a guideline on the management of mild, non-proteinuric hypertension in pregnancy.

Design  Simple, interrupted time series analysis.

Setting  Four maternity units in Scotland.

Population  Women delivering a live or stillborn baby.

Methods  Dissemination of the guideline under the auspices of a national clinical effectiveness programme, supported by a national launch meeting and feedback from a survey of obstetricians highlighting aspects of care that could be improved.

Main outcome measures  Appropriateness of initial investigation and subsequent clinical management, and costs of guideline development and implementation activities.

Data collection  Twenty-four months pre-intervention and 12 months post-intervention data were abstracted from a random sample of case notes.

Results  Initial investigation was consistent with recommendations for 59.9% out of 1263 women and subsequent clinical management for 67.6% out of 1081 in whom a diagnosis could be made from available data. There were no significant changes in the appropriateness of initial investigation (10.6%; 95% confidence interval [CI] −0.1% to 19.3%; decreasing by 1.2% per month post-implementation, 95% CI −2.5% to 0.1%) or clinical management (−0.3%; 95% CI −8.7% to 11.2%). Guideline development and implementation cost an estimated £2784 per maternity unit in Scotland.

Conclusions  Clinical care of mild hypertension in pregnancy remains highly inconsistent. The lack of the intervention effect may be related to the complexity of the guideline recommendations and the nature of the implementation strategy.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

Hypertensive disorders affect approximately 10% of pregnancies.1 There is confusion about optimal management of mild hypertension and its potential to progress to severe disease. An evidence-based clinical guideline was developed to promote more structured and efficient care, thereby enhancing clinical risk management.1

The use of valid guidelines can improve clinical practice, especially if accompanied by effective dissemination strategies.2 However, both the context within which guidelines are delivered and the nature of targeted clinical behaviours may also influence their uptake.3,4 The hypertension guideline was disseminated under the auspices of the Scottish Programme for Clinical Effectiveness in Reproductive Health (SPCERH), a professionally led initiative that brings together educational, audit and research activities relating to reproductive health care. Clinical effectiveness programmes have become an integral part of efforts to enhance interprofessional networking and improve the quality of care across several priority areas in Scotland.5 However, the impact and costs of support from such programmes upon guideline implementation have not been previously formally evaluated. We investigated the impact and costs of a strategy to support implementation of the guideline.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

The study was an uncontrolled interrupted time series analysis. This design can detect whether an intervention effect is significantly greater than the underlying (secular) trend.6 Such designs are appropriate in evaluating implementation strategies when it is difficult to randomise or identify a suitable control group—as with the national distribution of a guideline.

The study participants comprised maternity units in two teaching and two district general hospitals in Scotland. The units were also selected to balance geographical location and likelihood of compliance with the guideline, informed by an earlier survey of self-reported practice.7 The patient population consisted of women delivering a live or stillborn baby.

The guideline was disseminated nationally by SPCERH. Following a launch meeting attended by senior clinicians from all consultant-led obstetric units, copies were distributed to all obstetric consultants, specialist registrars, practising midwives and general practice principals in Scotland. The guideline was accompanied by feedback from a survey of obstetricians' reported practice highlighting aspects of care that could be improved.

Two main outcomes were derived from guideline recommendations.

  • 1
    Appropriate initial investigation by: confirmation of mild hypertension by checking two blood pressure measurements, and investigation of proteinuria, including screening for infection and re-checking ‘++’ proteinuria.
  • 2
    Appropriate clinical management during the subsequent week based on the care levels set out in Table 1.
Table 1.  Recommended packages of care according to the guideline, The Management of Mild, Non-proteinuric Hypertension in Pregnancy, and study criteria for appropriate management.
Care packageEligible womenComponents of care packageStudy criteria for appropriate care
Routine antenatal care‘Spot’ hypertension not sustained on assessment At least two out of three to be met:
• 1–2 documented contacts with professionals
• avoidance of admission (unless for labour)
• avoidance of investigations not indicated by guideline
Basic surveillanceConfirmed mild hypertensionBlood pressure recording and urine ‘dipstix’ twice weekly; clinical appraisal of fetal size and wellbeing; single estimate of serum urate, urea and electrolytes, full blood count, plateletsAt least three out of four:
• 2–3 documented contacts
• avoidance of admission
• avoidance of investigations not indicated by guideline
• appropriate anti-hypertensive treatment
Enhanced surveillanceDiastolic blood pressure sustained at over 100 mmHg or with mild hypertension and an incremental rise more than 25 mmHg since booking, clinical suspicion of poor fetal or maternal wellbeing or abnormal results on basic surveillance blood testsBlood pressure recording and urine dipstix at least 3 times per week; weekly serum urate, urea and electrolytes, full blood count, platelets, liver function tests; scan assessment of fetal size and liquor volume; cardiotocography assessment of fetal wellbeingAt least three out of four:
• 3–4 documented contacts
• avoidance of admission
• avoidance of investigations not indicated by guideline
• appropriate anti-hypertensive treatment
Specialist management (outwith guideline)Severe hypertension, proteinuria or abnormal findings during enhanced surveillanceReferral for specialist obstetric careDocumented referral to an obstetrician

A preliminary exploration of the data had indicated that the use of ‘strict’ outcome measures, such as requiring all four rather than three out of four criteria to be met for basic and enhanced surveillance, would have reduced compliance levels to near zero. Consequently, the ability of the study to detect any changes in practice would have been reduced. Therefore, ‘relaxed’ outcome measures were used to enhance sensitivity of the analysis.

The use of anti-hypertensive therapy was assessed from 20 weeks of gestation until delivery. Inappropriate care comprised the use of drug therapy after 32 weeks of gestation if diastolic blood pressure was less than 100 mmHg. Where drug therapy was considered justifiable, use of either labetolol or methyldopa achieved guideline compliance. The economic analysis was planned to measure the cost effectiveness of changes in practice for the main outcomes.

Trained data collectors screened case records to identify women with an episode of diastolic blood pressure of = 90 mmHg, proteinuria of ‘+’ or greater or both arising after 20 weeks of gestation. Data were abstracted for eligible cases using a standard form. Baseline data included the age of the women, gestation and parity at first visit. Data on management within the week of detection of raised diastolic blood pressure or proteinuria included: initial and subsequent diastolic blood pressure readings, urine dipstix testing and number of recorded clinical contacts, admissions (including recorded indications), documented review by an obstetrician, and investigations conducted (including documentation of abnormal findings). From 20 weeks of gestation onwards, data were collected on usage, timing and type of anti-hypertensive therapy and mode and outcome of delivery.

The guideline recommendations were based upon evidence of clinical rather than cost effectiveness. In some circumstances, the benefits gained from a change in clinical behaviour (e.g. reduced admissions) may be outweighed by the costs of interventions to promote that change in behaviour (e.g. educational meetings).8 Therefore, the economic evaluation was planned to incorporate data on the effects and costs of recommended clinical management as well as the effects and costs of guideline development and implementation. Subsequent modelling would determine whether the costs of development, implementation and subsequent management were worth any gains in health. For reasons that become apparent in the results, this analysis was not undertaken. It was also planned to present data on the costs of guideline development and implementation alongside data on the extent of any change in clinical behaviour; these data might then be transferable to other clinical contexts.4 For this, the costs of development and implementation were estimated from SPCERH documentary sources. Guideline development costs included the costs of literature search, development group meetings, travel, writing up and peer review. Implementation costs included those of the national launch (i.e. planning, venue, catering, and participant travel expenses), audit and feedback, and printing and distribution of the guideline. Staff costs were derived from standard sources.9–11 Methods used to derive estimates of the level of behaviour change are detailed below.

Given uncertainties in calculating sample sizes for time series analyses, the number of time points and the number of cases per time points were chosen largely on pragmatic grounds.12,13 Twenty-four monthly pre-intervention data sampling points were necessary to judge any preceding trends or seasonality (periodicity) in clinical management. Following a gap of eight months during which implementation activities occurred, 12 monthly post-intervention sampling points were used. Forty cases per month (10 per maternity unit) provided 80% power at 5% significance to estimate the proportion compliant at each time point to within 15%. This equated to a total planned sample size of 1440 cases. Random quota sampling using national records (SMR 02) produced 50 records to be screened for eligible cases per maternity unit per month (7200 in total).

Data were entered onto an Access Database. A logistic regression model that adjusted for clustering within each month was fitted. Robust standard errors were calculated using time period as a clustering effect. Estimates of level and post-implementation slope were calculated in terms of percentage change. As it was also possible that the guideline implementation strategy reduced variability (i.e. clinical practice became more uniform), Levene's test for equality of variances (pre to post) was used to identify any such effects.

The mean costs of guideline development and implementation per maternity unit in Scotland were calculated, based on SMR 02 returns. As the guideline was stated to be valid for a period of three years, costs were spread over this time.

The study was approved by the relevant local research ethics committees.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

We identified 1475 out of 7200 women (20.5%) with an episode of diastolic blodd pressure of =90 mmHg, proteinuria of ‘+’ or greater or both arising after 20 weeks of gestation. Detailed data were abstracted from a quota of up to 10 cases per maternity unit per month, giving 1263 cases for analysis.

The median age of women at booking was 28 [interquartile range (IQR) 24–31] with a median of zero previous completed pregnancies (IQR 0–1) and zero miscarriages and induced abortions (IQR 0–1). Median gestation at delivery was 40 weeks (IQR 38–40). A total of 433 (34.3%) of women had induced labours, of which 183 (14.5%) were recorded to be indicated by hypertension. A total of 285 (22.6%) of women underwent caesarean section, of which 60 (4.8%) were indicated by hypertension. There were no temporal trends in these characteristics over the study period.

Initial investigation was consistent with recommendations for 757 (59.9%) women (Table 2). Time series analysis showed no significant increase in the compliance level (10.6%; 95% confidence interval [CI] −0.1% to 19.3%; decreasing by 1.2% per month post-implementation, 95% CI −2.5% to 0.1%; Fig. 1a). There was no evidence of reduced variability in mean compliance (Levene's test: F= 0.351; P= 0.558), that is, clinical practice did not become more uniform following dissemination of the guideline.

Table 2.  Compliance with criteria for appropriate initial investigation.
OutcomesNo. eligibleNo. compliantPercentage compliance
  • *

    The total does not come to 1288 because 25 women experienced high diastolic blood pressure readings and proteinuria simultaneously.

Initial investigation of high diastolic blood pressure65338158.3
Initial investigation of proteinuria63539662.4
Overall compliance for initial investigation1263*75759.9
image

Figure 1. (a) Percentage compliance and 95% CI for initial investigation. (b) Percentage compliance and 95% CI for clinical management.

Download figure to PowerPoint

Subsequent clinical management was consistent with recommendations for 731 women (67.6%, based upon 1081 in whom a diagnosis could be made from available data; Table 3). Compliance was highest (81.3%) for 739 women requiring routine antenatal care; lack of compliance mainly being related to inappropriate investigation. Compliance was lowest (4.9%) in the 143 requiring enhanced surveillance, largely because of insufficient evidence of monitoring. For example, only 48 (33.6%) had recorded plasma urate, an indicator of fetal prognosis.14 However, in 139 cases (97.2%), anti-hypertensive treatment was initiated or withheld appropriately. There was no evidence of a change in level in the appropriateness of clinical management (−0.3%; 95% CI −8.7% to 11.2%; Fig. 1b) nor in the variability of the mean compliance (Levene's test: F= 2.169; P= 0.150).

Table 3.  Compliance with criteria for appropriate clinical management.
Criteria for recommendation levels of careNo. compliantPercentage compliance
Routine antenatal care (n= 739)
1 to 2 contacts within seven days64286.9
No inpatient admission or admission for labour62184.0
Avoidance of unnecessary investigations51069.0
At least two out of three criteria met60181.3
 
Basic surveillance (n= 43)
2 to 3 contacts within seven days2660.5
No inpatient admission or admission for labour3683.7
Appropriate investigations00
Appropriate anti-hypertensive treatment4195.3
At least three out of four criteria met2046.5
 
Enhanced surveillance (n= 143)
3 to 4 contacts within seven days4128.7
No inpatient admission or admission for labour9062.9
Appropriate investigations10.7
Appropriate anti-hypertensive treatment13997.2
At least three out of four criteria met2517.5
 
Specialist care required (n= 156)
Documented contact with obstetrician10366.0
Overall compliance (n= 1081)74969.3

The total cost of guideline development and implementation was estimated at £66,809, or £2784 per maternity unit in Scotland, implementation activities accounting for 60% of the cost. Given the absence of any significant effects, the economic evaluation took the form of a cost-minimisation analysis.15 This suggested that the development and implementation of the guideline was not cost effective.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

The care of pregnant women with raised blood pressure or proteinuria did not improve following guideline implementation by a national clinical effectiveness programme. Women with suspected or proven mild non-proteinuric hypertension continue to receive inconsistent clinical care. Nearly a third of women eligible for routine antenatal care were inappropriately investigated, while others with potentially more serious disease are often insufficiently monitored.

This study represents the first rigorous evaluation of a guideline implementation strategy led by a national clinical effectiveness programme. Methodological limitations need to be considered as potential explanations for our findings. Firstly, the outcome measures may have been insensitive to changes in practice. Nonetheless, ‘relaxed’ compliance criteria were applied to increase the sensitivity of findings. Secondly, as illustrated by the wide error around each data point (Fig. 1), this study lacked sufficient precision to detect modest changes in practice. Visual examination of the graph for initial investigation suggests a small increase in the intercept, raising the possibility that a small effect was not detected. There was, however, no indication of any similar change for clinical management. Thirdly, no data were collected over the intervention period and it is possible that a temporary effect, related to raised awareness of the guideline, was missed. In planning this study, we focussed resources available for data collection so as to detect any secular trends in the pre-intervention period and sustainable post-intervention effects.

Assuming that the design was sufficiently rigorous, the lack of any observed improvement in clinical practice might be related to the nature of the implementation strategy, the guideline or both. It is unclear why implementation strategies work in one context but not another. Similar studies indicate inconsistent effects of postal dissemination strategies and suggest that other activities are required to change practice.16,17 In a recent systematic review, multifaceted (combined) interventions did not appear more effective than single interventions.4 Furthermore, the effectiveness of more locally intensive and costly implementation activities (e.g. educational outreach) cannot be guaranteed.4

The guideline was introduced because of confusion among professionals over what constitutes optimal management. Mild hypertension in pregnancy presents a particular difficulty to obstetricians who need to balance calls for lower thresholds of clinical suspicion18 against others advocating less intrusive antenatal care.19 Although the type of focussed, structured clinical care advocated by the guideline, and optimal risk management need not be mutually exclusive, a postal survey indicated that obstetricians often found the guideline difficult to interpret.7 Therefore, the simplification of current guideline recommendations may offer one way forward towards improving the quality of care.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

No improvements in clinical care were detected following the dissemination of a national guideline under the auspices of a national clinical effectiveness programme, supported by a launch meeting and feedback of self-reported practice. The lack of the intervention effect may be related to the complexity of the guideline recommendations and the nature of the implementation strategy. Clinical care of mild hypertension in pregnancy is highly inconsistent and further action to promote improved quality of care is required.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References

The authors would like to thank the lead consultants and data collectors for each hospital as follows: Dr Nivisson Russell and Jennifer Comrie (Ayrshire Central Hospital); Dr Tahir Mahmood, Morag Telfer and Maureen Lamb (Forth Park Hospital); Dr Wang Liston, Dorothy Grooby, Lynne Glasgow, Nicola Clark and Fiona Elliot (Simpson Memorial Maternity Pavilion); and Frances MacLeod, Joan Ritchie and Jackie Heggarty (Aberdeen Maternity Hospital). Steven Simoens from HERU assisted with estimation of costs for the economic evaluation. RF was funded by a joint Medical Research Council/Chief Scientist Office Special Training Fellowship in Health Services Research. The Nuffield Trust funded the study. The Health Services Research Unit is core funded by the Scottish Executive Health Department. However, the views expressed are those of the authors and not necessarily those of the funding bodies.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. References
  • 1
    Scottish Obstetric Guideline and Audit Project. The Management of Mild, Non-Proteinuric Hypertension in Pregnancy. Scottish Project for Clinical Effectiveness in Reproductive Health, Edinburgh, 1997.
  • 2
    Grimshaw J, Russell I. Effect of clinical guidelines on medical practice: a systematic review of rigorous evaluations. Lancet 1993;342: 13171322.
  • 3
    Foy R, MacLennan G, Grimshaw J, Penney G, Campbell M, Grol R. Attributes of clinical recommendations that influence change in practice following audit and feedback. J Clin Epidemiol 2002;55: 717722.
  • 4
    Grimshaw JM, Thomas RE, MacLennan G, et al. Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technol Assess 2004;8(6).
  • 5
    Penney GC, Calder AA. The Scottish Programme for Clinical Effectiveness in Reproductive Health: lessons from a three-year programme. Health Bull 2000;58: 261266.
  • 6
    Shadish WR, Cook TD, Campbell DT. Quasi-experiments: interrupted time-series designs. Experimental and Quasi-Experimental Designs for Generalised Causal Inference. Boston: Houghton Mifflin, 2002171206.
  • 7
    Foy R, Penney G, Greer I. The impact of national clinical guidelines on obstetricians in Scotland. Health Bull 2001;59: 364372.
  • 8
    Mason J, Freemantle N, Nazareth I, Eccles M, Haines A, Drummond M. When is it cost effective to change the behaviour of health professionals? JAMA 2001;29882992.
  • 9
    Netten A, Dennett J, Knight J. Unit costs of health and social care 2001. Canterbury, Personal Social Services Research Unit, University of Kent, 2002.
  • 10
    Nursing and Midwifery Staffs Negotiating Council. Nursing and Midwifery Staffs Conditions of Service and Rates of Pay. Leeds: Department of Health, 2001.
  • 11
    Review Body on Doctors' and Dentists' Remuneration. Thirtieth Report. London: HMSO, 2001.
  • 12
    Crabtree BF, Ray SC, Schmidt PM, O'Conner PJ, Schmidt DD. The individual over time: time series applications in health care research. J Clin Epidemiol 1990;43: 241260.
  • 13
    Crosbie J. Interrupted time-series analysis with short series; why it is problematic; how it can be improved. In: GottmanJM, eidotrs. The Analysis of Change. New Jersey: Lawrence Erlbaum Associates, 1995: 361395.
  • 14
    Redman C, Beilin L, Bonnar J, Wilkinson R. Plasma-rate measurements in predicting fetal death in hypertensive pregnancy. Lancet 1976;13701373.
  • 15
    Drummond MF, O'Brien B, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes, 2nd edition. Oxford: Oxford University Press, 1997.
  • 16
    Mason J, Freemantle N, Browning G. Impact of Effective Health Care bulletin on treatment of persistent glue ear in children: time series analysis. BMJ 2001;323: 10961097.
  • 17
    Bloor K, Freemantle N, Khadjesari Z, Maynard A. Impact of NICE guidance on laparoscopic surgery for inguinal hernias: analysis of interrupted time series. BMJ 2003;326: 578.
  • 18
    Department of Health, Welsh Office, Scottish Office Department of Health, Department of Health and Social Services NI. Why mothers die. Report on Confidential Enquiries into Maternal Deaths in the United Kingdom 1994–1996. London: HMSO, 1998.
  • 19
    Department of Health. Changing childbirth. Report of the Expert Advisory Group. London, 1994.

Accepted 23 March 2004