Adolescent behavior in health care services utilization and study participation

(1)

DISSERTATIONS | VIRVE KEKKONEN | ADOLESCENT BEHAVIOR IN HEALTH CARE SERVICES... | No 332

uef.fi

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND Dissertations in Health Sciences

ISBN 978-952-61-2043-0 ISSN 1798-5706

Dissertations in Health Sciences

THE UNIVERSITY OF EASTERN FINLAND

VIRVE KEKKONEN

ADOLESCENT BEHAVIOR IN HEALTH CARE SERVICES UTILIZATION AND STUDY PARTICIPATION

This follow-up study aimed to investigate factors related to health care services utilization

and study attrition among adolescents and young adults. The study results indicate gender

differences, as well as psychosocial and socioeconomic problems among young people who are frequently using health care services.

Systematic selection bias caused by socio- demographic characteristics is possible in adolescent mental health research, and should be

investigated before causal conclusions a drawn from the research results.

VIRVE KEKKONEN

(2)

(3)

Adolescent behavior in health care services

utilization and study participation

(4)

(5)

VIRVE KEKKONEN

Adolescent behavior in health care services utilization and study participation

To be presented by permission of the Faculty of Health Sciences, University of Eastern Finland for public examination in University of Eastern Finland, Kuopio, on Friday, May 20th, 2016, at 12

noon

Publications of the University of Eastern Finland Dissertations in Health Sciences

Number 332

Institute of Clinical Medicine, School of Medicine, Faculty of Health Sciences, University of Eastern Finland

Kuopio 2016

(6)

Grano Oy Jyväskylä, 2016

Series Editors:

Professor Veli-Matti Kosma, M.D., Ph.D.

Institute of Clinical Medicine, Pathology Faculty of Health Sciences

Professor Hannele Turunen, Ph.D.

Department of Nursing Science Faculty of Health Sciences

Professor Olli Gröhn, Ph.D.

A.I. Virtanen Institute for Molecular Sciences Faculty of Health Sciences

Professor Kai Kaarniranta, M.D., Ph.D.

Institute of Clinical Medicine, Ophthalmology Faculty of Health Sciences

Lecturer Veli-Pekka Ranta, Ph.D. (pharmacy) School of Pharmacy

Faculty of Health Sciences

Distributor:

University of Eastern Finland Kuopio Campus Library

P.O. Box 1627 FI-70211 Kuopio, Finland http://www.uef.fi/kirjasto

ISBN (print): 978-952-61-2043-0 ISBN (pdf): 978-952-61-2044-7

ISSN (print): 1798-5706 ISSN (pdf): 1798-5714

ISSN-L: 1798-5706

(7)

Author’s address: Faculty of Health Sciences School of Medicine

University of Eastern Finland KUOPIO

FINLAND

Supervisors: Professor Eila Laukkanen, M.D., Ph.D.

School of Medicine

University of Eastern Finland KUOPIO

FINLAND

Professor Hannu Valtonen, Ph.D.

Department of Health and Social Management University of Eastern Finland

KUOPIO FINLAND

Professor Jukka Hintikka, M.D., Ph.D.

School of Medicine University of Tampere KUOPIO

FINLAND

Reviewers: Professor Pirjo Mäki, M.D., Ph.D.

Institute of Clinical Medicine University of Oulu

OULU FINLAND

Professor Pekka Rissanen, Ph.D.

School of Health Sciences University of Tampere TAMPERE

FINLAND

Opponent: Professor Riittakerttu Kaltiala-Heino, M.D., Ph.D.

School of Medicine University of Tampere TAMPERE

FINLAND

(8)

“Voi joutua itkemään vähän, jos on antanut kesyttää itsensä.”

(Antoine de Saint-Exupéry) Perheelleni

(9)

Kekkonen, Virve

Adolescent behavior in health care services utilization and study participation University of Eastern Finland, Faculty of Health Sciences

Publications of the University of Eastern Finland. Dissertations in Health Sciences 332. 2016. 70 p.

ISBN (print): 978-952-61-2043-0 ISBN (pdf): 978-952-61-2044-7 ISSN (print): 1798-5706 ISSN (pdf): 1798-5714 ISSN-L: 1798-5706

ABSTRACT

Adolescents are in biological and psychosocial transition and therefore at a vulnerable age for health concerns. Frequent utilization of health care services has been associated with female gender, mental health concerns, and problems at school. However, other factors related to frequent health care utilization among adolescents have been poorly explored.

Adolescent studies suffer from high attrition rates. Dropping out has been related to male gender and a lower socioeconomic status. Thus, little is known about the types of selection bias that may distort study results. This study aimed to investigate factors possibly related to health care utilization and study attrition among adolescents and young adults.

Students aged 13–18 years participated in a mental health questionnaire survey at their schools in 2005, and completed a postal questionnaire at the age of 18–23 years in 2011. The general population-based baseline sample consisted of 4171 (females 53.4%) participants, of whom 1827 (43.8% of baseline, females 61.4%) gave consent to be contacted for a follow-up, and 797 (50.3% of consented, females 70.9%) participated in the follow-up. Data on the utilization of health care services were gathered from the medical records of primary health care. Frequent health care utilization in 2005 (N = 793 participants, females 70.6%) and emergency room services utilization during 2005–2010 (N = 416 participants, females 67.3%) were investigated. Regression models with background factors as covariates were used to explain health care services utilization and the study participation.

Frequent utilization of health care services associated with female gender (OR 3.6, 95%

CI 1.8–7.3, P < 0.001), alcohol consumption (OR 3.2, 95% CI 1.3–7.5, P < 0.01) and mental health symptoms in females (Youth Self Report (YSR) subscale score withdrawn/depressed OR 1.9, 95% CI 1.1–3.4, P < 0.05), and self-reported somatic symptoms in males (YSR subscale score somatic symptoms OR 5.2, CI 95% 1.5–18.1, P < 0.01). The utilization of emergency room services was associated with female gender and psychosocial problems. In males, better school performance in the native language (IRR 0.1, 95% CI 0.0–0.5, P < 0.01) and in general subjects (IRR 0.3, 95% CI 0.1–0.8, P < 0.05) associated with fewer emergency room visits. Emergency services use associated with mental health problems in young adulthood in males (OR 10.5, 95% CI 1.9–59.6, P < 0.001) and alcohol consumption in teenage in females (OR 2.1, 95% CI 1.0–4.4, P < 0.05). Not working or studying (IRR 2.0, 95% CI 1.3–3.1, P < 0.01) and being single (IRR 2.6, 95% CI 1.6–4.0, P < 0.001) among females in young adulthood associated with a higher emergency room visit rate. An individual pattern measured as

(10)

unobserved heterogeneity associated with the utilization of emergency room services (OR 0.5, 95% CI 0.3–0.6 for males and OR 0.6, 95% CI 0.5–0.7 for females, P < 0.001). Study participation related to female gender (OR 2.3, 95% CI 1.9–2.7, P < 0.001) and symptoms of depression and anxiety (YSR subscale score anxious/depressed OR 1.1, 95% CI 1.0–1.1, P <

0.001). Adolescents with an older age (OR 0.9, 95% CI 0.9–1.0, P < 0.01), lower school performance (in the native language OR 0.8, 95% CI 0.5–1.2, P < 0.05, and in general subjects OR 0.7, 95% CI 0.6–0.9, P < 0.01), and with no hobbies (OR 0.5, 95% CI 0.4–0.7, P < 0.001) were less likely to participate in the follow-up phase.

The study results indicate gender differences, as well as psychosocial and socioeconomic problems among adolescents who are frequently using health care services. In addition, there might be an individual style in health care services utilization among young people.

Therefore, health care providers should invest in adolescents and young adults recurrently using health services by screening for psychosocial problems and providing health care services for them in multiple sources. Systematic selection bias caused by socio- demographic characteristics is possible in adolescent mental health research, and should be investigated before causal conclusions a drawn from the research results.

National Library of Medicine Classification: WA 330, WS 462, WS 463

Medical Subject Headings: Adolescent; Adolescent development; Emergency medical services; Follow-up studies; Health behavior; Health services/utilization; Mental health; Primary health care; Selection bias.

(11)

Kekkonen, Virve

Nuorten terveyspalveluiden käyttö ja tutkimukseen osallistuminen Itä-Suomen yliopisto, terveystieteiden tiedekunta

Publications of the University of Eastern Finland. Dissertations in Health Sciences 332. 2016. 70 s.

ISBN (print): 978-952-61-2043-0 ISBN (pdf): 978-952-61-2044-7 ISSN (print): 1798-5706 ISSN (pdf): 1798-5714 ISSN-L: 1798-5706

TIIVISTELMÄ

Terveyshuolet yleistyvät nuoruusiän biologisen ja psykososiaalisen kehitysvaiheen aikana.

Nuorten terveyspalveluiden käyttöä selittäviä yksilöllisiä taustatekijöitä on tutkittu melko vähän. Tutkimuksissa toistuva terveyspalveluiden käyttö on ollut yhteydessä naissukupuoleen, mielenterveysongelmiin ja koulunkäyntivaikeuksiin. Tutkimusten ongelmana on usein ollut tutkimusotoksen katoaminen seurannan aikana. Etenkin pojat ja alemman sosioekonomisen aseman omaavien perheiden lapset luopuvat herkemmin tutkimukseen osallistumisesta. Silti aineiston valikoitumiseen liittyvän harhan mahdollisuudesta on vain vähän tietoa. Tässä tutkimuksessa tarkasteltiin nuorten ja nuorten aikuisten terveyspalveluiden käyttöä sekä tutkimusotokseen valikoitumista mahdollisesti selittäviä taustatekijöitä.

Nuorten psyykkisen hyvinvoinnin seurantakyselytutkimus yleisväestölle toteutettiin kouluissa 13–18-vuotiaille opiskelijoille vuonna 2005 ja kirjeitse 18–23-vuotiaille vuonna 2011. Alkuvaiheeseen osallistui 4171 (tyttöjä 53,4 %) nuorta, joista 1827 (43,8 % alkuvaiheeseen osallistuneista, tyttöjä 61,4 %) antoi luvan seurantakyselyä varten ja 797 (50,3 % luvan antaneista, tyttöjä 70,9 %) osallistui lopulta seurantaan. Terveyskäyntitiedot kerättiin terveyskeskuksen sairauskertomusjärjestelmästä. Yhteensä 793 nuoren (tyttöjä 70,6 %) toistuvaa terveyspalveluiden käyttöä tutkittiin vuodelta 2005 ja 416 nuoren (tyttöjä 67,3 %) päivystyskäyntejä tutkittiin vuosilta 2005–2010. Nuorten yksilöllisten taustatekijöiden vaikutusta terveyspalveluiden käyttöön ja tutkimukseen osallistumiseen tarkasteltiin regressioanalyysimallien avulla.

Toistuvat terveyspalvelukäynnit olivat yhteydessä naissukupuoleen (OR 3.6, 95% CI 1.8–

7.3, P-arvo <0.001), tyttöjen nuoruusiän mielenterveysoireisiin (Youth Self Report (YSR) - alaluokka vetäytyminen/masentuneisuus OR 1.9, 95% CI 1.1–3.4 P-arvo <0.05) ja alkoholinkäyttöön (OR 3.2, 95% CI 1.3–7.5, P-arvo <0.01) sekä poikien fyysisiin oireisiin (YSR-alaluokka somaattiset oireet OR 5.2, CI 95% 1.5–18.1, P-arvo <0.01). Päivystyskäyntejä oli enemmän nuorilla naisilla ja niillä, joilla oli psykososiaalisia ongelmia. Nuorten miesten nuoruusiän hyvä koulumenestys äidinkielessä (IRR 0.1, 95% CI 0.0–0.5, P-arvo <0.01) ja yleisaineissa (IRR 0.3, 95% CI 0.1–0.8, P-arvo <0.05) vähensi päivystyskäyntejä. Sen sijaan nuorten miesten nuoren aikuisiän mielenterveysongelmiin (OR 10.5, 95% CI 1.9–59.6, P- arvo <0.001), nuorten naisten nuoruusiän alkoholinkäyttöön (OR 2.1, 95% CI 1.0–4.4, P-arvo

<0.05), nuoren aikuisiän syrjäytymiseen koulu- ja työelämästä (IRR 2.0, 95% CI 1.3–3.1, P-

(12)

arvo <0.01) sekä sinkkuna olemiseen (IRR 2.6, 95% CI 1.6–4.0, P-arvo <0.001) liittyi päivystyskäyntien määrän lisääntymistä. Yksilöllinen käyttäytymismalli oli yhteydessä etenkin naispuolisten tutkittavien päivystyskäynteihin (miehet OR 0.5, 95% CI 0.3–0.6 ja naiset OR 0.6, 95% CI 0.5–0.7, molempien P-arvo <0.001). Tutkimusotokseen valikoituivat useammin tytöt (OR 2.3, 95% CI 1.9–2.7, P-arvo <0.001) ja ne, joilla esiintyi ahdistus- ja masennusoireita (YSR-alaluokka ahdistuneisuus/masentuneisuus OR 1.1, 95% CI 1.0–1.1, P- arvo <0.001). Sen sijaan ne nuoret, jotka olivat iältään vanhempia (OR 0.9, 95% CI 0.9–1.0, P- arvo <0.01), joilla oli heikompi koulumenestys (äidinkielessä OR 0.8, 95% CI 0.5–1.2, P-arvo

<0.05 ja yleisaineissa OR 0.7, 95% CI 0.6–0.9, P-arvo <0.01) ja joilla ei ollut harrastuksia (OR 0.5, 95% CI 0.4–0.7, P-arvo <0.001), valikoituivat harvemmin seurantavaiheen tutkimusotokseen.

Toistuva terveyspalveluiden käyttö on jakautunut sukupuolen sekä psykososiaalisten ja sosioekonomisten taustatekijöiden suhteen. Todennäköisesti on myös olemassa yksilöllinen terveyspalveluiden käyttötyyli, jota ei voida selittää taustatekijöiden avulla. Toistuvasti terveyskeskuksessa käyvien nuorten psykososiaalisia ongelmia tulisi seuloa. Lisäksi nuorille ja nuorille aikuisille suunnattujen terveyspalveluiden saatavuuteen tulisi kiinnittää huomiota useilla eri palveluita tarjoavilla tahoilla. Tutkimusotokseen valikoituminen taustatekijöiden vuoksi voi aiheuttaa tilastollisen harhan, mikä vahvistaa aineiston edustavuuden analysoinnin tärkeyttä ennen johtopäätösten tekemistä.

Luokitus: WA 330, WS 462, WS 463

Yleinen suomalainen asiasanasto: akuuttihoito; mielenterveys; nuoret; perusterveydenhuolto; psyykkinen kehitys; seurantatutkimus; terveyskäyttäytyminen; terveyspalvelut.

(13)

Acknowledgements

My doctoral studies were accomplished within the Doctoral Program of Clinical Research, Faculty of Health Sciences, University of Eastern Finland (UEF). I want to express my thanks to all the funding sources. My dissertation work was financially supported by Valtion tutkimusrahoitus (VTR, previously EVO), and grants for my doctoral studies were provided by the Olvi Foundation, Suomen nuorisopsykiatrinen yhdistys ry, Psykiatrian tutkimussäätiö, Suomen lääketieteen säätiö, Päihdelääketieteen yhdistys ry, and OY H.

Lundbeck Ab.

I want to express my gratitude to all of my tutors for their guidance, advice, support and education. First, I must thank Prof. Eila Laukkanen, who has been an encouraging teacher and role model for me, in both research and clinical work as a psychiatrist. Secondly, I thank Prof. Hannu Valtonen, who led me into the world of the health care statistics and, most importantly, made statistics interesting and possible to learn. And thirdly, but not least, I want to thank Prof. Jukka Hintikka for specific instructions based on his clinical and scientific medical experience, especially in the revision of my articles.

My doctoral studies were an inevitable consequence of joining the research group of the Adolescent Psychiatry Department at Kuopio University Hospital (KUH). Therefore, I give my thanks and appreciation to the members of the research group. Special thanks to my doctoral student colleague, Petri Kivimäki B.M. His co-operation and efforts with the processing of the health care services data and altruistic help with my various problems regarding data and statistics were essential for this dissertation. Also, I thank my senior colleagues, Docent Tommi Tolmunen and Docent Soili Lehto, for pedantic scientific advice, precise criticism and support. Thanks to Noora Heikkinen for research assistance, Prof.

Ulrich Tacke, Marja-Liisa Rissanen M.Sc. and Outi Kaarre M.D. for their supportive professional comments. Furthermore, thanks to all my colleagues and co-workers who have encouraged me.

Thanks to Kuopio Primary Social and Health Care Management and Jari Saarinen for permission and the possibility to investigate the medical records of the participating adolescents. I also want to thank Tuomas Selander for statistical counseling and Prof. Heikki Tanila for comments on my adolescent brain development article. In addition, thanks to the Learning Center of UEF and the library of KUH for excellent services and help.

And finally, I assign most responsibility for my professional achievements to my closest family and friends. First, I must emphasize my dear parents Helena and Launo, who have shown me the meaning of true care, love, devotion, and hard work. They taught me that I must always believe in goodness and never give up. Moreover, I thank my dear siblings, sister Kirsi and little brother Simo. My courage and confidence must have been absorbed from them! I am so proud of our warm relations, and the great times we have spent together as a whole family. And of course, I give lots of thanks to my friends Viivi & Jarkko and Elina

(14)

& Antti. Unforgettable, hilarious excursions and serious talks with them have been a pleasure that I have always look forward to.

And most of all, I want to thank Ari, my cherished, my love. He has truly been there for me in good times and bad, and supported me in everything with his unconditional love and endless encouragement.

Kuopio, 16^th January 2016 Virve Kekkonen

(15)

List of original publications

This dissertation is based on the following original publications, which are referred to in the text by their Roman numerals:

I Kekkonen VK, Kivimäki P, Valtonen H, Tolmunen T, Lehto SM, Hintikka J, Laukkanen E. Psychosocial problems in adolescents associate with frequent health care use. Family Practice 32: 305-10, 2015.

II Kekkonen V, Valtonen H, Kivimäki P, Tolmunen T, Lehto S, Hintikka J, Kaarre O, Laukkanen E. Individual patterns as a determinant of the use of emergency room services among adolescents and young adults: a follow-up panel data analysis. Submitted manuscript.

III Kekkonen V, Kivimäki P, Valtonen H, Hintikka H, Tolmunen T, Lehto SM, Laukkanen E. Sample selection may bias the outcome of an adolescent mental health survey: results from a five-year follow-up of 4171 adolescents. Public Health 129: 162-172, 2015.

The original publications were adapted with the permission of the copyright holders.

(16)

(17)

List of abbreviations and definitions

A-DES Adolescent Dissociative Experiences Scale

ASEBA Achenbach System of Empirically Based Assessment AUDIT Alcohol Use Disorder Identification Test

BDI Beck Depression Inventory CI Confidence Interval ER Emergency Room

ICPC International Classification of Primary Care IRR Incidence Rate Ratio

LR Likelihood-Ratio Test OR Odds Ratio

RESET Ramsey Specification Test SD Standard Deviation

SPSS Statistical Package for the Social Sciences STATA Statistics and Data, Statistical software package

STM Sosiaali- ja terveysministeriö (Ministry of Social Affairs and Health) SVT Suomen virallinen tilasto (Official Statistics of Finland)

TAS-20 The twenty-item Toronto Alexithymia Scale

THL Terveyden ja hyvinvoinnin laitos (National Institute for Health and Welfare) VIF Variance Inflation Factor

YSR Youth Self Report Z-score Standard Score

(20)

(21)

Health behavior is largely adopted at a young age during a critical somatic and mental developmental period in life (Aalberg and Siimes 2007; Hanson and Chen 2007). Adolescent health behavior reflects individual and environmental characteristics and is lowered by psychosocial difficulties such as poverty and poor social relationships (Hanson and Chen 2007; Viner et al. 2012). Although adolescents in generally are a very healthy population, they are at a vulnerable age for health concerns, of which most are transient and related to puberty and psychosocial development (Buchanan, Eccles and Becker 1992; Campo et al.

1999; Rhee 2005; Aalberg and Siimes 2007). Of the severe health issues, the rates of mental health problems such as depression increase in puberty, and approximately one-fifth of adolescents suffer from a psychiatric disorder (Costello, Copeland and Angold 2011).

Adolescents account for a notable proportion of users of primary health care services (Mölläri and Saukkonen 2014). In addition, the utilization of emergency room services increases in adolescence and young adulthood (Callahan and Cooper 2010; Gnani et al.

2014). Furthermore, a large proportion of the resources of primary health care services are used by frequent attenders (Vedsted and Christensen 2005; Vila et al. 2012). Frequent primary health care utilization by adolescents has been related at least to problems at school and with mental health (Vila et al. 2012; Homlong, Rosvold and Haavet 2013). By comparison, frequent emergency room utilization by adolescents has been associated with female gender, mental health problems (Newton et al. 2010), socioeconomic deprivation (Newton et al. 2010; Rudge et al. 2013), a lower health status (Lau et al. 2014), and alcohol- related injuries (Linakis et al. 2009).

There are challenges in studying adolescents’ health behavior (Hinshaw et al. 2004;

Hooven et al. 2011). Representativeness of the data is one of the main issues in evaluating the significance of research findings. Nevertheless, dropping out is common in adolescent mental health research, and may distort the results (Hooven et al. 2011). However, very little is known about the types of systematic selection bias that may affect studies in follow-up settings.

There is an abundance of previous research on the health and behavior of adolescents.

However, adolescent health behaviors such as health care services utilization and study participation, as well as associated factors, have been poorly explored. Therefore, there has been a need to increase understanding of the patterns of adolescent health behavior.

Adolescent health behaviors were approached by investigating the utilization of health care services and research participation in a longitudinal general population-based study setting. The aims were to investigate 1) the frequent utilization of primary health care services among adolescents, and 2) the utilization of emergency room services among adolescents and young adults. Furthermore, the aim was to investigate 3) the study participation and representativeness of the study sample of adolescents at two study points:

a) in the informed consent phase and b) in the follow-up participation phase. This

(22)

dissertation provides a summary of three original research articles and their efforts to solve the presented questions concerning adolescent health behavior. The combination of follow- up data from the Adolescent Mental Health Survey and data from the local Primary Health Care Register enabled this research.

(23)

2 Review of the literature

2.1 ADOLESCENCE AND HEALTH

Adolescents generally form a very healthy population. Health behavior patterns are largely adopted at a young age in psychosocial and cultural contexts. However, development from childhood to adulthood is related to an increased risk of health complaints, of which mental health problems are often the most severe. In addition, psychosocial background factors have a significant role in the health development of adolescents.

2.1.1 Somatic and psychosocial development in adolescence

Adolescence (age 12–22 years) is an important developmental and maturational period between childhood and adulthood. Adolescents share similar developmental challenges as they disengage from their parents and increase individual autonomy with help from their peer relationships. Youth culture is a form of social cohesion. Together, adolescents have similar norms, values, behaviors, styles, and interests, which are purposely different from those among adults (Aalberg and Siimes 2007). Personal identity, behavior, and ambitions in life are established in late adolescence and early adulthood (Aalberg and Siimes 2007).

Development in adolescence is affected by individual, environmental, and cultural factors, and especially by the childhood family (THL, Lasten ja nuorten mielenterveys 2014).

Puberty changes the body image and sexuality, which are adopted during the physical, mental, and behavioral developmental phase in adolescence (Aalberg and Siimes 2007).

Rapid alterations in mood, behavior, and health complaints in adolescence are considered to be strongly related to hormonal changes in puberty and bodily maturation (Buchanan, Eccles and Becker 1992). Greater social activity and a stronger interest in the body and health among adolescent girls might be related to biological gender differences. Regardless, insecurity towards developmental bodily changes occurs even among the healthiest teenagers (Aalberg and Siimes 2007).

During puberty, healthy adolescents become more vulnerable to transient health concerns. Adolescent girls have more somatic complaints than boys (Poikolainen, Kanerva and Lönnqvist 1995). Headache, stomach ache, and sleeping problems are common symptoms among adolescents and often occur among the same individuals (Luntamo, Sourander and Aromaa 2015). The increase in somatic complaints in adolescence has been related to psychosocial problems such as mental health symptoms, substance use, and negative life events (Poikolainen, Kanerva and Lönnqvist 1995; Luntamo, Sourander and Aromaa 2015). Compared to the normal timing of puberty, the early or late occurrence of puberty has been associated with an increased prevalence of somatic symptoms, especially among girls. Of these somatic symptoms, headaches and musculoskeletal pains are the most common (Rhee et al. 2005). In addition, medically unexplained symptoms are common in adolescence, and are associated with female gender, a lower parental socioeconomic status, parental separation, and mood and behavior problems (Campo et al. 1999).

(24)

Compared to adults, adolescents need enhanced privacy and more intensive support from health care professionals to make decisions concerning health issues (Aalberg and Siimes 2007). Adolescents are already able to make most of the decisions concerning their own health, but they require family support in evaluating the need to seek help from health care services. Parents are still the primary decision makers for younger adolescents, while older adolescents are more autonomous (Ryan et al. 2011). Adolescents with health complaints need to be interviewed both alone and with their parents (Aalberg and Siimes 2007). Furthermore, it has also been recommended to investigate the psychosocial situation of adolescents with somatic complaints (Luntamo, Sourander and Aromaa 2015). Compared to older patients, access to a reliable and regular doctor might be more important to young patients in this transitional phase from childhood to adulthood (Ryan et al. 2011).

2.1.2 Mental health in adolescence One-fifth of all adolescents suffer from psychiatric disorders. The rates of mental health problems such as depression, panic disorder, agoraphobia, and substance use increase in puberty. Of the most serious mental health concerns, about 5% of adolescents have psychotic symptoms (Costello, Copeland and Angold 2011). Over half of mental health disorders in adults appear before the age of 14, and adolescents double the rates compared to mental health disorders as children (THL, Lasten ja nuorten mielenterveys 2014).

There are gender differences in adolescent mental health. For example, the incidence of depression is two times greater among adolescent females compared to adolescent males (Evans et al. 2005). Thus, social phobia may be a stronger predictor of depression among adolescent males compared to females (Väänänen et al. 2011). During adolescence in general, girls show more internalizing symptoms, while boys show more externalizing symptoms (Ormel et al. 2012). Conduct disorder is two or three times more common among adolescent boys than girls (Hipwell and Loeber 2006). Nevertheless, problem behavior is more serious in adolescent girls and predicts a range of adverse outcomes such as poor somatic and mental health, substance abuse and dependence, and antisocial behavior (Hipwell and Loeber 2006). Adolescents are in a vulnerable age for adverse life events and experiences. Psychological trauma is strongly related to an increased risk of anxiety disorders, depression, and substance use among adolescents (Haravuori, Suomalainen and Marttunen 2009). Mental health problems also continue over generations, as shown among offspring of depressed parents who have a 2–4 fold higher risk of depression compared to the offspring of non-depressed parents (Evans et al. 2005). In addition, first-degree relatives of schizophrenia patients have an approximately 10 times greater risk of schizophrenia compared to the general population (Gur et al. 2005).

Alcohol use increases in adolescence, and drinking is associated with a number of health risks and social problems (Skala and Walter 2013). Although the prevalence of adolescent alcohol use has decreased during the past few decades, binge drinking in particular is still very common among adolescents, and girls have become equal to boys (Raitasalo et al.

2012). Binge drinking in adolescence has been related to mental health and behavior problems, school problems, increased risky behavior, and parental mental health and substance use disorders (Brown et al. 2008; Niemelä 2010). Evidence is especially strong for

(25)

an association between external behavioral symptoms and substance use among adolescents. It is also possible that the early initiation of substance use may predict internalizing mental health symptoms in adulthood, especially among females (Miettunen et al. 2014). Alcohol-related psychosocial problems often occur among the same individuals and lead to social exclusion (Kekkonen, Kivimäki and Laukkanen 2014). Alcohol consumption appears to be more common among adolescents from single-parent families and adolescents with abundant allowances (Ahlström and Karvonen 2010). The Finnish drinking culture has been considered to be very permissive (Härkönen and Österberg 2010).

Adolescents taste alcohol in the presence of their family, but their real alcohol consumption occurs with peers (Foltran et al. 2011). Among adults, alcohol is mainly consumed at home, and an alarming proportion of children are affected by their parents’ alcohol use and may consider it a frightening experience (Raitasalo 2010).

Mortality among Finnish young people is low (SVT, Kuolemansyyt 2014). Due to higher morbidity related to suicides, accidents, and violence, males are overrepresented in the death statistics of young people (SVT, Kuolemansyyt 2014). However, during the past decades, there has been an alarming increase in suicide rates and more violent suicide methods among females (Lahti et al. 2011). Among deceased 15–24-year-olds, over one- third have died by committing suicide. Therefore, suicide is a common cause of death among young people, especially among males, who commit three-quarters of all suicides (SVT, Kuolemansyyt 2014). Psychopathology is the most significant risk factor for both suicides and suicide attempts among adolescents (Hendin et al. 2005). Furthermore, suicide risk increases in relation to the number of psychiatric disorders, especially comorbid affect disorders and substance use disorders, among adolescents (Hendin et al. 2005).

Unfortunately, adolescents who commit suicide have experienced severe problems for several years (Hendin et al. 2005). However, less than half of young suicide victims have received psychiatric care (Pelkonen and Marttunen 2003).

Mental health disorders are closely related to restricted performance in studies and working life. Difficulties in school performance, a short educational career, and remaining without an education are related to mental health problems in adolescence (Ostamo et al.

2007). For example, depression among adolescents and young adults is related to an increased rate of somatic health concerns and days off from school when compared to their non-depressed peers (Haarasilta 2003).

Mental health disorder is the most common reason for receiving disability allowance in young adults (Ahola et al. 2014). In addition, a long-lasting and severe history of mental health illness is related to disability allowance among young adults (Ahola et al. 2014).

Compared to young people with other reasons leading to disability, a major proportion of young adults with mental health disorder have experienced many psychosocial difficulties in their childhood, such as parental divorce, parental alcohol problems, bullying, learning disabilities, severe somatic disorders, and experiences of physical or sexual abuse (Ahola et al. 2014).

(26)

2.1.3 Health behavior patterns and psychosocial factors in adolescence The concept of health capital, “good health” as commodity, assumes that individual health depreciates with age and can be increased by investments such as education and health care (Grossman 1972). Health behavior can be defined as personal actions to maintain good health and as a reflection of personal health beliefs (Mosby 2009). There are differences in individual health behavior among adolescents. Individual health behavior patterns are largely adopted at a young age, concurrently with the development of personality and lifestyle from a wider perspective. Individual health behavior patterns in adolescence may also be reflections of youth culture.

The milieu in which adolescents grow up has an important role in supporting and enhancing the development of normative health behavior (THL, Lasten ja nuorten mielenterveys 2014). The adolescent health status is lowered by poor wealth, income inequality, lower educational possibilities, an adverse growth milieu, and poor family and peer relationships (Viner et al. 2012). Therefore, psychosocial factors such as psychological and somatic development, social affairs, and experiences in childhood and adolescence form a basis for health and welfare in adulthood (THL, Lasten ja nuorten mielenterveys 2014).

Differences in socioeconomic status and psychosocial circumstances are closely related to health. Environmental background factors such as material circumstances are crucial in defining individual health and predicted health behavior. According to Finnish statistics, compared to lower socioeconomic groups, adults in higher socioeconomic groups follow dietary recommendations, exercise more often, and live longer and healthier lives (Roos et al. 2007). The parental socioeconomic status has an effect on factors associating with the socioeconomic status and educational level of adolescents (Lahelma et al. 2007). A lower socioeconomic status has been related to poorer daily health behaviors such as diet, exercise, and cigarette smoking among adolescents (Hanson and Chen 2007). Students in vocational schools drink excessive amounts of alcohol and are cigarette smokers more often compared to students in high schools (Helakorpi et al. 2007; Laaksonen et al. 2007). Furthermore, the peer relationships of adolescents might have a significant role in adopting weight-related behavior such as regular exercise and fast food consumption (Ali, Amialchuk and Heiland 2011). Regardless, daily health behaviors are mostly adopted and modeled from caregivers, whereas substance use is learned in peer relationships (Hanson and Chen 2007).

The concept of social exclusion has not been stabilized, but it has been used to describe a fall into chronic poor economic and social circumstances by an individual or whole family (SVT, Tulonjakotilasto 2009). The risk factors and manifestations of social exclusion appear to continue over generations from parents to their offspring. Dropping out from school, unemployment, chronic poverty, and health problems often occur concurrently, and are associated with social exclusion (Mohajer and Earnest 2010).

(27)

2.2 HEALTH CARE SERVICES UTILIZATION AMONG ADOLESCENTS (I- II)

After puberty, personal growth and development highlights individual differences in health behavior, health care services needs and utilization. It is well known that psychosocial difficulties are closely related to an increased rate of health complaints among adolescents.

Nevertheless, there is less knowledge of the increased utilization of health care services and associated factors.

2.2.1 Health care services and utilization Health care services are divided into primary, secondary, and tertiary health care. Here, focus is on primary health care, which refers to the municipally arranged monitoring of the health of the population with various services. Primary health care services are provided at municipal health centers (STM, Terveyskeskukset 2014). In addition, emergency room services offer walk-in medical treatment for patients with acute unexpected health complaints. Adolescents account for a notable proportion of patients in primary health care (Mölläri and Saukkonen 2014). Furthermore, the utilization of emergency room increases in adolescence and young adulthood (Callahan and Cooper 2010; Gnani et al. 2014).

It is well known that the utilization of health care services is determined by both demand (such as health problems, but also socio-demographic and psychological factors) and supply factors (availability and organization of various health services). Furthermore, the differences in the utilization of health care services are associated with demand side factors such as individual characteristics, health needs, and socio-demographic factors.

Frequent health care attenders account for a large proportion of contacts with primary health care. It has been suggested that the top 10% of individuals with frequent health care utilization account for up to 50% of all health care visits (Vedsted and Christensen 2005).

There is no standard definition for frequent health care utilization. For example, an integer threshold from 2 to 24 contacts in 12 to 24 months has often used to describe frequent attendance (Vedsted and Christensen 2005). Frequent attenders in primary health care more often suffer from somatic, psychiatric, and social problems compared to non-frequent health care attenders (Smits et al. 2009).

There are gender differences in health, health behavior, and health care utilization. In general, women are healthier than men, but they use a higher proportion of health care resources (Nguyen et al. 2011). Mortality is higher among males than females (Nguyen et al. 2011), and gender differences in mortality increase in adolescence (Gissler et al. 2006). In childhood, boys have more health problems and health care services utilization compared to girls (Gissler et al. 2006). After puberty, the utilization rate of health care services increases among adolescent girls (Gissler et al. 2006; Nguyen et al. 2011), and remains higher through life among women compared to men (Nguyen et al. 2011). Higher rates of utilization of health care services among females can be partly explained by gynecological reasons (Gissler et al 2006). Furthermore, there might be social differences and differences in behavioral style between genders in the utilization of health care.

(28)

Socioeconomic inequalities are strongly related to health care utilization in the Finnish population. People with a high income use more occupational health and private practice services, while visits to municipal health care services are more common among people with a low income (Manderbacka et al. 2007).

There are also individual patterns in health behavior and the utilization of health care services. Individual health behaviors are a more complicated issue than only defining the assortment of individual characteristics. The differences between individuals can be seen as variation in an individual’s behavior over time and variation in behavior between individuals (Gunasekara et al. 2014). Here, the concept of individual patterns refers to differences between individuals, or behavioral patterns that cannot be explained by known background characteristics (i.e., represent “unobserved heterogeneity”). Individuals may differ in their personal load of stress, adaptive regulation, physiological state, and external disturbances (Yashin et al. 2008). The non-medical reasons affecting care-seeking behavior can be divided in two groups: Firstly, there are differences between population groups in their care-seeking behavior, e.g. between males and females or persons of different ages.

These differences are observable in the same sense that they can be statistically explained by gender and age, reflecting e.g. cultural differences between population groups. Secondly, there are differences between individuals that cannot be statistically explained in a given dataset (“unobserved heterogeneity”). People with the same gender and age might have an individual pattern of behavior, and the existence and degree of this heterogeneity can be observed in panel data analysis, but not the factors behind it. For example, hidden heterogeneity in age-related characteristics may cause incorrect causal conclusions concerning the associated factors (Yashin et al. 2008).

2.2.2 Frequent primary health care services utilization among adolescents Among adolescents, the most common medical reasons for attending primary health care include respiratory tract problems, gastrointestinal problems, and signs or symptoms of diseases such as a poor appetite, aches, and pains (Vila et al. 2012). Psychosocial difficulties are related to an increased rate of somatic symptoms such as headache, stomach ache, and sleeping problems among adolescents in comprehensive school (Luntamo, Sourander and Aromaa 2015). Among frequent health care using adolescents, typical health complaints include upper respiratory tract infections, asthma, injuries, and acne (Kramer et al. 1997).

Few studies have investigated frequent health care use in adolescents. Among British secondary school pupils, 30% of adolescents were frequent attenders of primary healthcare, and they had 4 or more visits per year (Vila et al. 2012). Having been admitted to hospital, absence from school, current medical illness, and a previous need for psychiatric consultation were associated with frequent primary care use among adolescents (Vila et al.

2012). Moreover, in a study on Norwegian students (aged 15–16 years), dropping out from school was associated with frequent student health care use and referrals to mental health care (Homlong, Rosvold and Haavet 2013).

To our knowledge, however, there have been no studies on adolescent substance use or risky health behaviors in relation to frequent primary health care use. According to a Finnish health survey of young adults, mental symptoms such as depression increased the use of

(29)

health care services, and health care use was more prevalent among young women compared to men (Kestilä et al. 2007). Furthermore, survey results suggested that several childhood and current adverse experiences and high alcohol consumption were associated with psychological distress. In addition, psychological distress was associated with frequent use of health care services, but only few of distressed young adults had sought professional help for a mental health problem (Kestilä et al. 2007). Among adult patients, chronic diseases, mental health problems, and social problems have been associated with frequent primary health care visits (Smits et al. 2009).

2.2.3 Frequent emergency room services utilization among adolescents

The utilization of emergency room services increases in adolescence and young adulthood (Callahan and Cooper 2010; Gnani et al. 2014). Common reasons leading to this increasing use in young people include musculoskeletal issues, injuries, and respiratory tract infections (Gnani et al. 2014). Concerning mental health issues, the most common conditions leading to the need for emergency room services are depression, conduct disorders, substance use, and unspecified neurotic disorders (Mahajan et al. 2009). Injury-related emergency visits increase in puberty (Downing and Rudge 2006). In a Finnish emergency room survey, up to two-thirds of underage patients attending emergency room due injuries were under the influence of alcohol (Karjalainen et al. 2013).

Recurrent emergency room visits by adolescents are associated with female gender, older age, mental health problems (Newton et al. 2010; Burnett-Zeigler et al. 2012), socioeconomic deprivation (Newton et al. 2010; Rudge et al. 2013), a poorer health status (Lau et al. 2014), and alcohol-related injuries (Linakis et al. 2009). In a pediatric emergency room study, positive responses to a suicide-screening questionnaire associated with repeated emergency room visits in children (aged 8–12 years), and psychiatric hospitalization in adolescents (aged 13–18 years) (Ballard et al. 2013). A few American studies have revealed a relationship between higher rates of emergency visits and a lack of private health care insurance among adolescents and young adults (Callahan and Cooper 2010; Lau et al. 2014). In a pediatric emergency room study, a higher proportion of emergency room visits compared to all other visits to health care services associated with lower educational and income levels in the family and public insurance (Kroner, Hoffmann and Brousseau 2010).

In comparison to adolescents, among adult patients, frequent emergency room visits have been associated with an older age, chronic illness, psychological distress and mental health problems, socioeconomic distress such as a low educational level and low income, and a high level of use of other health care resources (Sun, Burstin and Brennan 2003; Jelinek et al. 2008; Stockbridge, Wilson and Pagán 2014). In addition, among adults, problem drinking of alcohol and illicit drug use have been associated with emergency room use more often than with other primary health care use (Cherpitel and Ye 2008; Jelinek et al. 2008).

The immediacy of the objective need in a medical sense for medical care is expected to determine the decision to use health care services. However, factors other than those related to the medical condition influence the decision to seek help from health care services such as emergency room services. Health care utilization might be affected by unobserved measures or individual characteristics and personal factors (Woolridge 2002). It might also

(30)

be possible that the utilization of health care services could be predicted by individual past behavior. Published research on individual health behavioral patterns and associations with health care utilization and emergency room utilization in adolescence or young adulthood was not found.

2.3 REPRESENTATIVENESS AND ATTRITION OF ADOLESCENT MENTAL HEALTH STUDIES (III)

Successful recruitment is essential for a representative study sample. Research drop-out rates are high in mental health studies concerning adolescents, possibly due to their developmental age. Adolescent mental health recruitment studies were summarized to provide information on attrition and possible sample selection among adolescents.

2.3.1 Study representativeness and bias

Study participation is a challenge in every study. Sample selection may distort study results and cause bias. Biased research results may induce biased causal conclusions and interpretations. Teenagers can be a difficult study population due to their developmental age. Some teenagers refuse to take part in surveys, and there is a risk of systematic selection associated with mental and somatic health and health care use. Attrition among teenagers has various causes related to both developmental and environmental factors, such as adolescent risk behavior and the lack of parental motivation (Hinshaw et al. 2004; Hooven et al. 2011). Furthermore, some adolescents might avoid revealing their private life issues, especially if it presumes consent from their parents.

Successful recruitment is essential for a representative study sample and reliable study results. The recruitment of adolescents in longitudinal studies concerning mental health is difficult and drop-out rates often exceed 50% (Hinshaw et al. 2004). Even so, in some previous general population-based follow-up studies in Finland, attrition rates among adolescents have been less than 30-40%, as in the Northern Finland Birth Cohort Study 1986 (Miettunen et al. 2014; Mäki et al. 2014) and in the Adolescent Mental Health Cohort Study (Kaltiala-Heino, Fröjd and Marttunen 2010; Väänänen et al. 2011),or even less than 10% in the Northern Finland 1966 Birth Cohort study (Riala 2004).

It has been concluded that the study procedure and recruitment methods must be well planned in advance to achieve the most representative study sample (Hooven et al. 2011).

Missing data have been noticed and investigated in some adolescent mental health studies by performing attrition or drop-out analyses. According to Finnish general population- based studies on adolescents, dropping out has been associated with male gender, an older age, a low parental educational status, and not living with both biological parents or living in other than intact families (Hurtig et al. 2005; Kaltiala-Heino, Fröjd and Marttunen 2010;

Väänänen, Fröjd and Ranta 2011; Mäki et al. 2014). Of the mental health aspects, adolescents with depression and with a parental history of psychiatric disorder have been found more likely to drop out of mental health studies concerning adolescents (Väänänen, Fröjd and Ranta 2011; Miettunen et al. 2014; Mäki et al. 2014). In addition, in a study investigating

(31)

adolescent depression and bullying, adolescents who had dropped out from the follow-up more often reported being bullies (Kaltiala-Heino, Fröjd and Marttunen 2010).

There might be similarities among adolescent and adult populations and factors associated with attrition. Thus, following the high drop-out rate in this follow-up study, the aim was to investigate the previous literature concerning attrition and sample selection, and the reliability of study results, especially among adolescents.

2.3.2 Recruitment studies on adolescent mental health To review the literature concerning recruitment and attrition studies on adolescent mental health, a PubMed search was conducted in May 2015 for studies published since 2002. Only papers in English were included. Study participation, mental health, recruitment, attrition, adolescence, retention, and consent were used as search terms. Papers were selected based on their abstracts. Altogether, eleven studies describing study recruitment difficulties were found (Table 1).

As shown in Table 1, the study protocol leading to the lowest rates of attrition was the recruitment of individuals through referrals from other health care systems (May et al. 2007;

Jaffee et al. 2009). Furthermore, incentives were offered in most of these studies (Audrain et al. 2002; Boys et al. 2003; Diaz and Pérez 2009; Jaffee et al 2009; Wilcox et al. 2012).

Compensation significantly increased compliance in two studies concerning drug use (Díaz and Pérez 2009; Wilcox et al. 2012). Recruitment was most unsuccessful among anorexia nervosa patients (Mc Dermott et al. 2004; Hewell, Hoste and Le Grange 2006). A lower parental educational level, a living environment including multiple deprivation, and non- Caucasian ethnicity reduced study participation (Audrain et al. 2002; de Winter et al. 2005;

May et al. 2007; Goodman and Gatward 2008).

Concerning study sample representativeness, six of the above studies described the possible selection bias with regard to study questions related to mental health issues or socioeconomic factors (Audrain et al. 2002; Boys et al. 2003; de Winter et al. 2005; May et al.

2007; Goodman and Gatward 2008; Fröjd, Kaltiala-Heino and Marttunen 2011). One of these studies evaluated both recruitment methods and factors associated with attrition (May et al.

2007). A lower educational level of the parents reduced their willingness to provide consent for the participation of the adolescents (Audrain et al. 2002).

Moreover, school performance among non-responding adolescents aged 15–16 years was lower in comparison to responders (Boys et al. 2003). Compared with responding children, non-responding children aged 10–12 years also had more problems at school and parents with a lower educational level, but no association was found between adolescent psychopathology and individual characteristics (de Winter et al. 2005). A lower response rate was also reported among children and adolescents (aged 15–16 years) who belonged to families with a low socioeconomic status (Goodman and Gatward 2008). In a Finnish longitudinal study among ninth-grade students (Fröjd, Kaltiala-Heino and Marttunen 2011), a lower parental educational status and lower school performance of adolescents were associated with high attrition rates. However, depression was not systematically associated with attrition.

(32)

The rest of these studies concentrated on various methods for improving adolescent recruitment and avoiding sample attrition, such as sharing recruitment information with medical facilities, health care providers, or the juvenile justice system, advertising in newspapers or on the radio, printing flyers, or providing incentives or other compensation (McDermott et al. 2004; Hewell, Hoste and Le Grange 2006; Díaz and Pérez 2009; Jaffee et al. 2009; Wilcox et al. 2012).

These studies indicate that the recruitment methods, potential incentives, and psychosocial characteristics of participants might have had an effect on study sample selection and attrition. In addition, consenting children, adolescents, or parents may differ from those who refuse participation. Therefore, it is possible that the factors associated with attrition might lead to a distorted study sample and biased results. The association between alcohol or illegal drug use and participation were not examined in these papers.

Adolescent behavior in health care services utilization and study participation

uef.fi

Dissertations in Health Sciences

VIRVE KEKKONEN

ADOLESCENT BEHAVIOR IN HEALTH CARE SERVICES UTILIZATION AND STUDY PARTICIPATION

VIRVE KEKKONEN

Adolescent behavior in health care services

utilization and study participation

Adolescent behavior in health care services utilization and study participation

Acknowledgements

List of original publications

Table of contents

List of abbreviations and definitions

2 Review of the literature