High Health Service Use Research Articles

Mental Disorders Among Frequent Attenders in Primary Care

Frequent attenders account for a large proportion of primary care (PC) contacts, referrals, and prescriptions. Psychosocial and emotional distress is related to the high use of health services. Few studies have focused on the association between mental disorders assessed using structured interviews and frequent use of PC services.The aim of this study was to determine the factors associated with frequent attendance at primary healthcare units, focusing specifically on mental disorders. A two-phase screening epidemiological study comparing frequent attenders and routine attenders in five primary health care units was designed. Three hundred eighteen frequent attenders and 203 patients who attended the same units on a routine basis were compared. Sociodemographic and clinical data were obtained from statistical records and medical charts. Patients with a total score equal or higher than 7 points on the General Health Questionnaire-28 (GHQ-28) were interviewed using the Schedules for Clinical Assessment in Neuropsychiatry. All the scores obtained on the GHQ were statistically different in the two populations. Frequency of mental disorders also differed significantly between both groups, with somatoform and affective disorders being the most prevalent ICD-10 categories among frequent attenders. The presence of depressive disorders and somatoform disorders is the most powerful predictive factor for frequent attendance. High comorbidity was found among frequent attenders with somatoform disorder. Frequent attendance at primary healthcare units is associated with depressive and somatoform disorders. Psychiatric comorbidity could be a confounder, particularly because affective and somatoform disorders often overlap in PC patients.

Importancia de los problemas reumáticos en la población de Cataluña: prevalencia y repercusión en la salud percibida, restricción de actividades y utilización de recursos sanitarios

ObjectiveTo determine the importance of chronic musculoskeletal problems in the adult population of Catalonia (Spain) and their effect on self-perceived health, activity restriction and use of health services. MethodsA population-based survey of 15,926 adults was performed. Multistage stratified sampling was performed. The variables gathered were sociodemographic characteristics, self-reported chronic health problems, self-perceived health, activity restriction and use of health services. Musculoskeletal problems were grouped into four categories: osteoarthritis-arthritis or rheumatism (OA), chronic dorsal or lumbar pain (LBP), chronic cervical pain (UBP), and osteoporosis. ResultsChronic health problems were reported by 77.4% of the adult population. The most frequent health problem was LBP, followed by UBP and OA. After adjustment by age was performed, female sex increased the risk of reporting OA, LBP, UBP and osteoporosis (OR=2.6, 1.5, 2.3, and 5.3, respectively). The prevalence increased with greater age and with lower socioeconomic status. After adjustment was performed by age, sex, social class and obesity, self-perceived health was worse in people with these problems (42.7% vs 11%). The four categories were the main causes of activity restriction in the last year (OR 2.70) and the last 15 days (OR=2.32) and were associated with a higher use of health services. ConclusionesLos problemas reumáticos son los principales problemas de salud crónicos declarados por la población adulta. La prevalencia es mayor es las mujeres, aumenta con la edad y en las clases desfavorecidas. Hay una asociación significativa entre declarar problemas musculoesqueléticos y salud autopercibida mala o regular, y mayor restricción de actividades y uso de servicios sanitarios.

Unsettled infant behaviour and health service use: A cross‐sectional community survey in Melbourne, Australia

To investigate factors associated with health service use by women and their infants in Victoria, Australia. Cross-sectional screening survey of 875 women with 4-month-old infants attending immunisation clinics in five local government areas in Melbourne between May 2007 and August 2008. The self-report instrument assessed socio-demographic characteristics, unsettled infant behaviour, maternal mood (Edinburgh Postnatal Depression Scale) and, the outcome, health service use during the first 4 months post-partum. Mothers and their infants used on average 2.8 different health services in the first 4 months post-partum (range 0-8). After adjustment for other factors, high health service use (defined as >3 different services) was more common in mothers whose infants were unsettled with persistent crying, resistance to soothing and poor sleep. A one-point increase on the unsettled infant behaviour measure was associated with an 8% (2-14%) increase in the use of >3 services, 9% (3-16%) in use of emergency departments, 7% (2-13%) in use of telephone helplines and 9% (3-14%) of parenting services. Poorer maternal mental health was also implicated with a one-point increase on the Edinburgh Postnatal Depression Scale associated with a 4% (0.4-8%) increase in the likelihood of using more than three services. Unsettled infant behaviour is associated with increased use of multiple health services. The high use of emergency departments by families with unsettled infants found in this study suggests that enhancement of primary health-care capacity might be required.

Characteristics of a Mild Head Injury Subgroup With Extreme, Persisting Distress on the Rivermead Postconcussion Symptoms Questionnaire

Kirsch NL, de Leon MB, Maio RF, Millis SR, Tan-Schriner CU, Frederiksen S. Characteristics of a mild head injury subgroup with extreme, persisting distress on the Rivermead Postconcussion Symptoms Questionnaire. Objective To examine baseline variables and identify characteristics of participants with extremely high reports of symptoms (ie, outliers) 12 months after mild head injury (MHI). Design A prospective cohort study of MHI with and without loss of consciousness (LOC) and/or posttraumatic amnesia (PTA) recruited from and interviewed at the emergency department (ED), with a follow-up telephone interview at 12 months. Setting Level II community hospital ED. Participants Participants (n=58) with MHI and LOC less than or equal to 30 minutes and/or PTA less than 24 hours and participants (n=173) with MHI but no PTA/LOC. Inclusion criteria: age greater than or equal to 18 years, less than or equal to 24 hours after injury, Glasgow Coma Scale score greater than or equal to 13, and discharge from the ED. Fourteen (6%) participants had extremely high scores on the Rivermead Postconcussion Symptoms Questionnaire (RPQ). Main Outcome Measures RPQ and questions on health services use and litigation. Results Characterizing the outlier cases are prior head injury, preinjury disability, history of substance use, unemployment, and elevated somatic symptoms at the ED. At 12 months, outliers had higher use of health services and litigation. Conclusions The existence of a subgroup with a distinctive pattern of baseline characteristics in combination with elevated somatic symptoms at the time of presentation to the ED suggests that further taxonomic distinctions may be warranted for the MHI population, each requiring appropriately targeted interventions for addressing symptomatic complaints.

Perceived Need for Mental Health Care and Service Use Among Adults in Western Europe: Results of the ESEMeD Project

Perceived Need for Mental Health Care and Service Use Among Adults in Western Europe: Results of the ESEMeD Project

Antisocial personality disorder: Co‐morbid Axis I mental disorders and health service use among a national household population

AbstractAntisocial personality disorder is highly co‐morbid with other mental disorders, particularly substance use disorders, but little is known about the use of mental health services among this group. The study investigated the use of mental health services and associations with psychiatric co‐morbidity among persons with a diagnosis of antisocial personality disorder within a representative UK household population. Weighted data from 8 395 respondents included 245 (2.9%) with a diagnosis of antisocial personality disorder. This diagnosis resulted in independent risks of substance abuse disorders, suicide attempts and generalized anxiety disorders when compared with the general population. Help‐seeking behaviour was more prevalent among the antisocial group than the general population but was primarily mediated through co‐morbid Axis I mental disorders and was not because of antisocial symptoms or other co‐morbid personality disorders. Among those with antisocial personality disorder, approximately one‐quarter had consulted a mental health professional in the year before interview. Of those, nearly 70% suffered from a co‐morbid anxiety disorder. High use of mental health services among persons with antisocial personality disorder appears to be a characteristic of a specific subgroup with high levels of anxiety. Copyright © 2009 John Wiley & Sons, Ltd.

Risk for Suicide among Medicaid Beneficiaries

Demographic, diagnostic, and service expenditure characteristics of Florida Medicaid enrollees who died by suicide were investigated. Among persons receiving Medicaid and Supplemental Security Income (SSI), findings indicate the most powerful predictors of suicide were involuntary psychiatric examination, mental health hospitalization, and high mental health service use. Among Medicaid enrollees not receiving SSI, strongest suicide predictors were mental health hospitalization, high expenditures for physical health medications, and involuntary psychiatric examination. Findings suggest reducing involuntary psychiatric examinations and mental health hospitalizations while improving physical health may reduce suicide in the Medicaid population. Comprehensive hospital discharge planning, adherence monitoring with follow-up care, training mental health providers in assessing suicide lethality, and providing adequate assessment time are all crucial to achieve these objectives.

Australian Women Seeking Counseling Have Higher Use Of Health Services

Despite a high prevalence of psychological distress and poor mental health in the Australian community, use of counseling services is very low. There has been only limited research examining the profile of those who do access counseling services, mainly in terms of demographic and health behavior variables. To extend our understanding of those who currently access counseling services, this study aimed to examine the broader pattern of health service utilization by women who consulted counselors, psychologists, or social workers in the past year compared with those who did not among a population-based sample of middle-aged Australian women, and to determine whether health service utilization was independently associated with use of counseling services, controlling for other known predictors. The cross-sectional population-based mail survey data for this study came from the third survey of the mid-aged cohort of the Australian Longitudinal Study on Women's Health, conducted in 2001. The sample comprised 11,201 women aged 50-55. The main study variable was a question asking whether they had consulted a counselor/psychologist/social worker in the past year. Only 6.9% of women had consulted a counselor/psychologist/social worker in the past year. After controlling for self-reported mental health status, health behaviors and demographic variables in multivariate analysis, consulting a counselor/psychologist/social worker in the past year was significantly and positively associated with consultations with general practitioners (> or =5 consultations; odds ratio [OR], 4.14; 95% confidence interval [CI], 2.35-7.27; p < .0001), specialist (> or =3 consultations; OR, 2.09; 95% CI, 1.66-2.63; p < .0001), and hospital doctor (OR, 1.35; 95% CI, 1.10-1.66; p = .004). Use of counseling services was not associated with use of other allied and complementary health services in multivariate analyses. Further research is needed to determine whether the strong independent link between self-reported use of counseling and other medical and health services among middle-aged women is best explained by general practice referral patterns, availability of services, economic factors, or different help-seeking patterns among women.

SEAL ET AL. RESPOND

We appreciate Rona’s concerns about population-based mental health screening of Iraq and Afghanistan veterans presenting to Department of Veterans Affairs (VA) medical facilities. The study population included veterans presenting to all outpatient clinics at 1 VA facility, not just primary care. Mental health screening may not have been a clinical priority during medical subspecialty visits, which may partially explain lower screening rates. Recently, in response to changes in national VA policy, postdeployment mental health screening rates in primary care have increased to more than 90%. Further, screening instruments used by the VA, specifically the primary care posttraumatic stress disorder (PTSD) screen, have demonstrated validity,1 and brief screens for PTSD symptoms have been used successfully to detect cases of PTSD in primary care.2 When population-based postdeployment mental health screens were first introduced in VA facilities in 2004, the extent of mental problems stemming from the conflicts in Iraq and Afghanistan were just surfacing. Hoge et al. described stigma surrounding mental illness as a barrier to treatment for soldiers, highlighting that soldiers in greater distress were less likely to seek help.3 Combat veterans might be more willing to disclose and accept treatment for mental health problems from the VA rather than the military. In addition, veterans may report mental health concerns months after returning home when at the VA, rather than immediately on their return when first screened by the military. We concur that universal screening should not be conducted if there is inadequate follow-through of positive screens. In our study, 73% of combat veterans with positive screens attended mental health appointments, as opposed to 18% of veterans not screened. Unfortunately, the majority of mental health visits occurred more than 90 days after the positive screens. Since this study was conducted, however, the VA has greatly augmented its mental health capacity, hiring nearly 100 new psychologists.4 Further, the VA has implemented a national model of integrated, collocated care in which mental health providers embedded in primary care conduct immediate assessment and triage of veterans who screen positive for mental health symptoms.4 Preliminary data from one integrated clinic for Iraq and Afghanistan veterans demonstrated high use of mental health services within primary care. We cannot wait for a randomized controlled trial, as Rona suggests, to decide whether we should continue to screen veterans for combat-related mental disorders. If we don’t ask, they may not tell, and we cannot afford the potential consequences of undetected mental illness in combat veterans.

The Association of Psychiatric Diagnoses, Health Service Use, and Expenditures in Children with Obesity-related Health Conditions

To examine the association of psychiatric diagnoses and use of health care services in children with obesity-related health conditions. A retrospective, longitudinal design was used to examine Medicaid claims data. The data set consisted of 13,688 youth diagnosed with type 2 diabetes, metabolic syndrome, dyslipidemia, or obesity. The presence of any type of psychiatric diagnosis was associated with higher health service use. In particular, the presence of an internalizing diagnosis was more consistently associated with higher service use than the presence of an externalizing diagnosis. Children with both an externalizing and internalizing disorder diagnosis had greater service use than children with a diagnosis in only one of these categories. These data highlight a subgroup of children with obesity-related health conditions who are at greater risk for higher health service use, and the need for further research on the association between psychiatric diagnosis and health service use.

Utility of the concept of minor depressive disorder: Evidence from a large rural community sample in a developing country setting

Background Evidence on the utility of the minor depressive disorder diagnostic construct in rural communities and developing countries is scarce. Aim To assess the utility and validity of definitive minor depressive disorder in Ethiopia by determining its impact and sociodemographic correlates. Methods Assessment of 68,378 adults, aged 15–49 years, living in a largely rural district in Ethiopia using the Composite International Diagnostic Interview. Results The lifetime prevalence of minor depressive disorder was 2.2% (95% CI = 2.1%, 2.3%). Age, marital status, education and unemployment independently predicted minor depressive disorder. Over 80% of cases used health services, 55.1% experienced persistent thoughts of death and 14.6% attempted suicide. Limitation Findings are based on lifetime estimates. Conclusions Minor depressive disorder is an important public health problem in rural Ethiopia, as shown by the associated high health service use and risk behaviour. Sociodemographic correlates suggest aetiological continuity with major depression. Thus our findings extend the clinical and public health utility of this diagnostic construct to rural community and developing country settings.

Evaluating the effectiveness of a women's crisis house: a prospective observational study

Aims and MethodTo investigate the effectiveness of the Women's Service crisis house in Croydon we performed an observational study prospectively measuring functioning, symptom severity and unmet needs before and after admission. Use of mainstream mental health services was also measured.ResultsWomen using the service had high use of mainstream mental health services, with 137 out of 269 (51%) requiring admission to a mainstream acute ward in the 4 years studied. The service was effective with an improvement in Global Assessment of Functioning scores from a median of 48 on admission to 67 on discharge (P&lt;0.001).Clinical ImplicationsThe women's crisis house was effective in providing for women who required high levels of mainstream mental health services, suggesting that it provides a valuable alternative to standard in-patient care.

History of domestic violence and health service use among mid‐aged Australian women

Objectives: To examine associations between history of domestic violence and health service use among mid‐aged Australian women, adjusting for physical and psychological health status and demographic factors.Methods: Population‐based cross‐sectional postal survey (1996) of the Australian Longitudinal Study on Women's Health. Of 28,000 women randomly selected, 14,100 (53.5%) aged 45–50 years participated. Logistic regressions were used to assess associations between domestic violence and health service use.Results: After adjusting for demographic variables, multivariate analysis revealed associations between ever having experienced domestic violence and three or more consultations in the previous 12 months with a family doctor (OR 2.07, 95% CI 1.68–2.55), hospital doctor (OR 1.77, 95% CI 1.44–2.17), or specialist doctor (OR 1.54, 95% CI 1.35–1.75), or being hospitalised (OR 1.36, 95% CI 1.20–1.54). After adjusting for demographic variables and physical and psychological health status, these associations were attenuated: three or more consultations with family doctor (OR 1.36, 95% CI 1.09–1.70), hospital doctor (OR 1.16, 95% CI 0.92–1.45), or specialist doctor (OR 1.14, 95% CI 0.98–1.32), and being hospitalised (OR 1.10, 95% CI 0.96–1.26).Conclusions: Physical and psychological status accounted for the associations between domestic violence and higher health service use, with the exception of GP consultations, which remained associated with domestic violence.Implications: Physical health status only partially explains the increased health service use associated with domestic violence, while both physical and psychological health status explained higher usage of specialist and hospital services. It seems likely that women who have experienced domestic violence may be seeking consultations from GPs for reasons additional to health status.

Placement Stability and Mental Health Costs for Children in Foster Care

Although prior population-based studies have found that children in foster care use more mental health services than their Medicaid peers, less is known about how different experiences in foster care impact the likelihood of mental health service use. The primary aim of this study is to test the hypothesis that instability of foster care placements is associated with higher costs for mental health care services. The secondary aim is to test the hypothesis that foster care children are also more likely to generate high costs for mental health services if they generate higher costs for non-mental health claims. Using administrative child welfare data linked to Medicaid claims, we assembled a unique retrospective cohort of adjudicated dependent children >2 years old who entered foster care between July 1993 and June 1995, spent at least 9 months in care, and were Medicaid eligible during a 1-year follow-up period. The primary outcome was high mental health service use, defined as having costs in the top decile of the sample. The primary independent variables were the number of foster care placements during the year and whether placements were interrupted by a return home for at least 1 month during that year (episodic foster care). We used logistic regression to estimate the association between placements and service utilization, with adjustment for age and physical health care costs. Of the 1635 children in the study, 41% had > or = 3 foster care placements, and 5% had episodic foster care during the year of observation. The top 10% of mental health service users accounted for 83% of the 2.4 million dollars in mental health costs. Both multiple placements and episodic foster care increased the predicted probability of high mental health service use. Higher physical health care costs also increased the probability of high mental health use for all children, but this increased probability was most dramatic among children with episodic foster care (probability of high mental health use: 0.78; 95% confidence interval: 0.42-0.94). Foster care placement instability was associated with increased mental health costs during the first year in foster care, particularly among children with increasing general health care costs. These findings highlight the importance of interventions that address the global health of children in foster care and may permit better targeting of health care resources to subgroups of children most likely to use services.

Introducing coordinated care (2): evaluation of design features and implementation processes implications for a preferred health system reform model

The study investigated why the goals of the Australian Coordinated Care trials for clients with complex care needs were not achieved. Significantly higher health service use and costs were incurred in the absence of clear evidence of improved client health outcomes. The validity of assumptions underpinning trial design and the success of implementation at each step in application of the model were examined. There were failures in both design and implementation. Many clients did not require care coordination. The funds pooling arrangements contributed to limited possibilities for service substitution and training of GP care coordinators was inadequate. Trial design did not focus on either clinical guidelines or consumer empowerment. Furthermore, the expectations of the overall national trial were unrealistic both in trial design and expected outcomes given the rigidities and realities of the Australian health care system. Broader system reform in the form of funds pooling and health services planning at the regional level, based on large populations, may be a more effective means to address problems of care coordination and an inflexible supply system.

Cost of outpatient care in patients with inflammatory bowel disease in a German University Hospital.

Because of its long duration, inflammatory bowel disease (IBD) causes high use of health services and high lifetime costs for medical care. The aim of the present study was to measure the costs of outpatient care in patients with IBD in a German University Hospital and to identify potentially relevant determinants of costs. The use of resources of 599 outpatient patients treated at a German University Hospital (65% Crohn's disease [CD] and 26% ulcerative colitis [UC]) was measured using a routine database. Costs of medical services (diagnostics and treatment) were considered as well as costs of medication. Resource use was valued using fee schedules for hospital services and pharmacy prices for drugs. The mean cost of one outpatient visit was Euros 162, including physician costs, laboratory costs, and costs of diagnostic procedures following the visit. For a subgroup of 272 patients, the mean annual cost for outpatient care was Euros 3171. Medication accounted for 85% of the total annual costs. Potential determinants, such as main diagnosis (CD or UC), sex, age, localization of disease, and occurrence of anemia, had no influence on costs, whereas complications of IBD and use of corticosteroids showed an impact on annual costs. This is the first time that the structure and range of outpatient treatment costs for IBD have been demonstrated for a German hospital.

Home treatment of ‘malaria’ in children in rural Gambia is uncommon

Home treatment with antimalarials is a common practice in many countries, and may save lives by ensuring that more malaria cases receive prompt treatment. Through retrospective surveys we found that home treatment of young children with antimalarials was uncommon in rural Gambia. Few families kept medicines in the home in case of illness, 28% kept paracetamol and only 8% kept chloroquine. Less than 10% of cases of childhood 'malaria' had been treated with chloroquine at home, and 69% of those giving home medication did not know the correct dosage for a child. The most common course of treatment was the use of paracetamol and/or tepid sponging to reduce fever, before the child was taken to a government health facility. Treating a child with antimalarials at home was more costly than other forms of treatment. The low cost associated with the use of health services for children and the limited availability of antimalarials outside major towns contribute to the high use of government health services. This shows that that home treatment cannot be assumed to be the predominant mode of malaria treatment throughout Africa, and highlights the need for country-specific policies based on accurate local knowledge of treatment practices in both rural and urban areas.

Factors associated with high use of public mental health services by persons with borderline personality disorder.

The research presented here was a pilot study to identify clinical factors associated with high use (as opposed to lower use) of inpatient psychiatric services by persons with borderline personality disorder. The initial sample was a random sample of English- and Spanish-speaking persons aged 18 to 60 years who had received at least one outpatient mental health service in the previous 90-day period and were enrolled in one of the participating mental health centers in King County, Washington. A random sample of persons who met selection criteria was randomly drawn; persons with high levels of use were oversampled to ensure adequate representation. Twenty-nine participants met full criteria for borderline personality disorder on the Personality Disorders Examination structured interview and completed all measures. Fifteen (52 percent) of these had a high level of use of inpatient services, and 14 did not. High use of inpatient psychiatric services was predicted by a history of parasuicide in the previous two years but not by the number or severity of parasuicides; by the presence and number of anxiety disorders but not by depression or psychotic or substance use disorders; and by poorer cognitive functioning. Life stressors, global functioning, and health service variables did not differentiate patients with high levels of service use from other patients with borderline personality disorder. Further research should explore these predictors of service use to determine whether they are replicated in larger samples, and treatments that target these variables should be evaluated.

Prevalence, incidence and stability of premenstrual dysphoric disorder in the community.

Despite an abundance of clinical research on premenstrual and menstrual symptoms, few epidemiological data provide estimates of the prevalence, incidence, co-morbidity, stability and correlates of premenstrual dysphoric disorder (PMDD) in the community. To describe the prevalence, incidence, 12 co-morbidity factors and correlates of threshold and subthreshold PMDD in a community sample of young women. Findings are based on prospective-longitudinal community survey of 1488 women aged 14-24, who were followed-up over a period of 48 months (follow-up, N = 1,251) as part of the EDSP sample. Diagnostic assessments were based on the Composite International Diagnostic Interview (CIDI) and its 12-month PMDD diagnostic module administered by clinical interviewers. Diagnoses were calculated using DSM-IV algorithms, but daily ratings of symptoms, as required, were not available. The baseline 12-month prevalence of DSM-IV PMDD was 5.8%. Application of the diagnostic exclusion rules with regard to concurrent major depression and dysthymia decreased the rate only slightly (5.3%). An additional 18.6 % were 'near-threshold' cases, mostly because they failed to meet the mandatory impairment criterion. Over the follow-up period only few new PMDD cases were observed: cumulative lifetime incidence was 7.4%. PMDD syndrome was stable across 48 months with < 10% complete remissions among baseline PMDD cases. The 12-month and lifetime co-morbidity rates were high (anxiety disorders 47.4%, mood disorders 22.9%; somatoform 28.4%), only 26.5 % had no other mental disorder. Particularly high odds ratios were found with nicotine dependence and PTSD. In terms of correlates increased rates of 4-weeks impairment days, high use of general health and mental health services, and increased rates of suicide attempts were found. In this sample of adolescents and young adults, premenstrual symptoms were widespread. However, DSM-IV PMDD was considerably less prevalent. PMDD is a relatively stable and impairing condition, with high rates of health service utilization, increased suicidality and substantial co-morbidity.

The use of primary, secondary, community and social care by families who frequently consult their general practitioner.

There have been very few studies that have examined either the use of other health and social services by patients who frequently consult their general practitioner, or the patterns of service use of frequently attending families. This study has three aims: to quantify the number of contacts to primary, secondary, community and social services of families who frequently attend their general practitioner; to examine the temporal patterns of these contacts and the patterns of use within families; and to determine the nature of these contacts and how they were valued by patients and health professionals. Thirty-five individuals in seven families completed diaries of health service use for 8 weeks; records from primary and community care records were analysed; health professionals and patients provided satisfaction scores for their contacts. The data were compared to standard datasets. The subjects made far more consultations with all health services than predicted, and less than half of these were with the general practitioner. Thirty-two per cent of the contacts with staff based within the health centre were not reported on the diary sheets. Temporal patterns of consulting were apparent within the families. The "index frequent attenders" within each family had most contacts. The differences between professional groups for the nature of the contact were marked, with over two-thirds of general practitioner contacts being for "specific symptoms or problems", and over two-thirds of health visitor contacts being for "weight problems". These findings show that patients who make high use of general practitioners' services also have a large number of contacts for health reasons with a range of other health and social services. These need to be considered as an outcome measure for trials of intervention for frequent attendance. More work is needed to explore why some individuals and their families make high use of health services.